Cities, Networks, and Globalization IV

 Defining Globalization Through Cities

I acknowledge for this section the limited purpose of describing globalization in relation to the role of cities as interregional connectors, particularly in light of a network conception of the city in topological space.  With this frame of reference in mind, globalization simply constitutes one of a multiplicity of descriptions of long distance, multidirectional network connections between city regions existing over the course of human history.  I will proceed from the position, mirroring those of Pred (1977), Bourne (1975) and, more recently, Taylor (2004), that such connections between city regions are continuously produced and reproduced over time. 

To isolate a particular impression on the importance of interregional commerce to the development of cities, Jacobs (1970: 35; 1984: 135-155) insists on the impossibility of urban economic development in the absence of commercial exchange processes between cities, resulting in the articulation, development, transformation, collapse, and reproduction of interregional economic systems anchored by cities.  The unique forms, in which such interregional city systems become manifest, demonstrate novelty in relation to the scale of the exchange networks developed.  That is to say, trade networks among Mediterranean cities in the Twelfth century manifest a geographically different form and scale of distances from contemporary supply chains for computer hardware integrating cities at opposite corners of Pacific Rim. 

The unifying characteristic of both such interregional systems of exchange remains the centrality of connecting cities as a condition of their existence.  Sources of differentiation can be linked explicitly to characteristics of the networks articulated in different interregional systems linking particular city regions at particular moments in time.  Especially, the types of transmission processes involved in distinct interregional networks yields an important source of difference.  The extent to which technologies permit, over time, more radical space-time compression enables more pronounced extensions of networks through Cartesian space, making interregional networks relatively more global in nature.  For example, information transmissions in the Twelfth century, involving particular methods and tools for written codification and transmission over relatively long distances between parties sharing capacities for translation (e.g. common written languages or easily translatable concepts in multiple written lingual structures), reflect particular constraints on the geographic extension of information transmission less apparent at the present time.  By contrast, contemporary methods of wireless information communication enhance the speed of point-to-point information transmission processes, facilitating significance extensions in the distances between connected stationary information producing and/or collecting/disseminating nodes. 

Continuing to emphasize the importance of integrating network theory into this understanding of globalization through cities, extensions of the length of networks never imply saturation of Cartesian space.  Globalization, as the extension of interregional networks over longer and longer distances, never implies that transmissions from networked processes occupy broad, continuous fields in Cartesian space or, even, the pathways between individual nodes.  To interject a Jacobsean image of interregional commerce anchored on cities, Twelfth century commodities flowed between Constantinople and Venice, but not to all points in between unless the transmission process, per se, has failed (e.g. interdiction of maritime transportation through piracy).  Today, information travels from Seattle to Bangkok, but not to the Northwestern Pacific coastal or Thai countrysides or to all other intervening points not otherwise connected to the media by which information is transmitted.  Wireless communications media do not perfectly or evenly saturate the global landscape with possibilities for connectivity, even with the appropriate accessing technologies – dead zones exist where no wireless communications media can reach.  In short, if I analyze globalization through the frame of reference of networks, then the effects of transmissions generated by processes associated with globalization cannot be theorized as universal, in the sense that they diffuse to touch all points in Cartesian space.  Rather, globalization describes the manifestation of effects, transmitted over long distances across diverse pathways, at discrete, identifiable nodes.  The imagery of geographically uneven development conveyed by this network theory of globalization underscores the relative importance of cities as spaces through which long distance transmissions emanate and from which information, other use values, and people become unevenly dispersed across city regions.    

Accepting these insights as a point of departure for an understanding of globalization consistent with Jacobs’ theorization of interregional city systems, this project seeks to move further to argue that interregional network connections are never simply commercial or, more generally, economic in nature.  Rather, the sort of networked processes that connect cities to form interregional city systems of the kind envisioned by Jacobs include cultural and political processes as well.  For example, movements of human beings, resulting in new diasporic settlement patterns for particular ethnic groups, may extend the geographic range for diverse collective cultural practices defining the ethnic group.  Such cultural process transmissions by means of human migration produce archipelagos of diasporic communities, constituted by structures of ethnic practices, extending out from their geographic bases.  For example, the migration of Mexican workers from the southern state of Oaxaca, moving through the region of Ciudad Juarez to the region of Chicago, generates an ethnic diasporic cultural network with nodes in the ethnic home region (Oaxaca), one transitional city region (Ciudad Juarez), and a terminal city region (Chicago), connected by transmission processes.  Interregional cultural networks of this kind reflect constant transformation as diasporic groups interact with the culture of “host” groups, as new members of the diaspora arrive from the geographic base, and as other means of information transmission enable processes at various nodes in the network to be transformed by the effects of processes at other points (e.g. changes in religious practices in a diasporic node effecting transformations at other diasporic nodes and in the geographic base nodes in the home country).

Political processes introduce other sources of interregional network development.  Such processes include the communication of legal or extra-legal norms across space and, importantly, their enforcement through the exercise of authority in myriad local contexts.  For example, the passage of statutory law by some state organization, claiming legislative authority over a jurisdiction with fixed boundaries containing multiple city regions, constitutes a given political process whose effects must be transmitted across jurisdictional space.  In accordance with the theory heretofore developed, the transmission of the law over space will be effected through the space of cities, with the technological means to undertake the required communication of information across regional spaces.  The speed at which information of the law can be received across jurisdictional space varies in relation to differential access to means of communication – other cities will receive the information first and diffuse it across their regions via appropriate means of communication.  From here the effects of the political process by which the law was passed will vary based on particular way in which other, local processes shape the enforcement process undertaken differentially across jurisdictional space.  In this manner, the interregional political network configured by the transmission of legal norms connects with intra-regional political networks in a continuous, multi-directional interchange of information and effects, shaping the enforcement of the law and the processes through which legal norms are produced at spatio-temporally defined and delimited network nodes. 

At this point, it is worth summarizing the arguments that I have, in this section, attempted to convey with regard to globalization.  First, it conforms to an imagery of multidirectional interregional network connections emanating from cities, which occupy a central role in the articulation of intra-regional networks integrating their non-urban hinterlands.  This conception of globalization is, thus, constructivist in the sense that it is constituted as an aggregation of relatively long networks that are, individually, constructed from the linking of individual, local nodes.[18]  It is not a monolithic force existing continuously in an omnipresent “out there,” antagonistically confronting individual, local agents and rendering them passive victims of globally extensive economistic logics.   Second, the imagery of globalization largely conforms to Jacobs’ understanding that cities actively constitute, as a matter of definition, interregional city systems.  On the other hand, I have sought to broaden the understanding conveyed by Jacobs to include not only networks constituted by commercial exchange processes but also by every other economic, cultural, and political process.  Cities, thus, exist as connectors for every kind of process articulated through networks over relatively long distances. 

Again, approaching from the opposite direction, every kind of process constituting a network over relatively long distances in Cartesian space requires a city to forge its connections.  Globalization, within this framework of networks, simply constitutes a concept for describing the complex range of interregional networks articulated through the space of cities, stretching ever-longer distances by means of transmission technologies that bend Cartesian space, creating the myriad topological spaces of networks.[19]  By this definition of globalization, every city is engaged in globalization to the extent that, by definition, every city articulates long distance connections with other cities. 

This constitutes the basic conception of globalization that I mean to convey within this project.  On the other hand, if I was to leave the discussion of globalization in this section at the idea of long distance interregional network articulations through cities, then I would be exposing this project to the impact of a glaring analytical absence: notably, the fact that cities are, generally, contained by political entities called nation-states.  In this regard, I have to specify not only the relationship of the city, as I have defined it so far, to the nation-state, but also the relationship of the nation-state to my conception of globalization.

            Nation-states, as they will be understood in this project, are organizations/entities constituted by structures of political processes whose spatio-temporalities can be patterned as networks, defined across a Cartesian space with given jurisdictional boundaries.  Such a Cartesian space can be described as relational insofar as the capacity of a nation-state to claim political sovereignty over a space depends on its relationship to the political processes defining other nation-states, contestation of boundaries (e.g. military actions) between nation-states, and other social processes (e.g. commercial exchange, immigration of culturally heterogeneous groups) transcending but penetrating its boundaries.  It also depends on myriad economic, cultural, and ecological processes internal to its boundaries, whose effects either strengthen or undermine the capacity of a nation-state to maintain its integrity.  In this sense, cultural processes defining a collective history and a common sense of nationhood, linked, in part, to the political process of citizenship, operate over the Cartesian space internal to the nation-state’s political defined boundaries as networks, linked structurally (as mutually constituted/constitutive sets of processes) and spatio-temporally (as overlapping and intersecting networks) to the political processes constituting the nation-state as source of legal authority and a sovereign within its space. 

Thus, as a nation-state, France exists as an overlapping set of political networks.  Such networks articulate the transmission of effects of state political processes by the central government in Paris to all of the nation’s geographic departments and the effects of diverse enforcement processes from the departments through concentric, hierarchical layers of authority to the central government in a multidirectional framework of determinations.  The capacity of the French state to maintain political control over a geographic space like Corsica depends, moreover, on the particular networks, facilitating the transmission of use values, people, and information to undertake both domestic processes of law enforcement and public services, and to undertake military operations to assert the political power of the French state relative to neighboring nation states.  In addition, cultural processes, consistent with Althusser’s (1971: 141-148) understanding of ideological state apparatus, occurring in innumerable spaces (e.g. schools, churches, households, cafes, etc…) articulate distinct networks through which information moves with diverse media (e.g. person-to-person conversation, books, television), reinforcing and/or undermining the constitutional legitimacy of the French state and the claim of French central state authority across innumerable geographical spaces within the defined boundaries of the nation-state.  Finally, patterns of commerce within local departmental regions and between departments and the Paris city region, shaped, in part, by the political process of the French central state, by those of the European Union, and by those of the World Trade Organization (e.g. concerning the continued existence of agricultural subsidization), articulate particular economic networks that may both reinforce and undermine the internal integrity of France as a nation-state in diverse ways. 

            An understanding of the nation-state as a particular kind of network space, encompassing articulated political, economic, and cultural networks, puts the nation-state on a footing with my network definition of the city.  The interaction of the processes articulating these networks over Cartesian space produces, with varying degrees of territorial integrity, an envelope in Cartesian space, encompassing a defined and variously legitimated or contested political boundary and a continuous range of internal space.  If the city exists as an intersection of local/intra-regional networks and long distance/interregional networks, then some of the networks intersecting in the city define it and its region as entities contained within the envelope of a nation-state.  For example, the networks constituted by federal political processes, certain cultural processes relating to a shared national history, and economic processes relating to internal/national exchange relationships bind the city region of Boston to the United States as a discrete geographic area contained by the political boundaries of the United States as a nation-state.

Critically, however, the nation-state is only one of myriad possible envelopes, articulated by networks, within which the city and its region may be contained.  Numerous contemporary accounts on the city and city regions, like those of the regional economic theory and the control and coordination theory approaches noted in the preceding chapter, reinforce conceptions of globalization as a set of economic processes promoting, at least in part, the disarticulation of cities from national economic space.  The notion of a “global commodity chain,” for example, expresses the conception of integrated production, circulation, and exchange networks for use values transcending national boundaries and, at least in part, undermining the capacity of individual nation-states to exercise effective macroeconomic policy with respect to the operation of commodity chains transcending and penetrating their boundaries (Gereffi et al., 1994).  In the same respect, Taylor (1994; 2004: 179-193, 196-214) presents an especially radical but compelling account, linked specifically to Jacobs’ rejection of the nation-state as a legitimate, coherent economic entity.  Specifically, Jacobs argues, nation-states constitute “grab bags of very different economies” (Jacobs, 1984: 32), combining politically captive but economically dynamic city regions with relatively stagnant city regions and wide, contiguous, but relatively disconnected, non-urban spaces separated from any city region.  For Jacobs, “cities are unique in their ability to shape and reshape the economies of other settlements” (Ibid: 32) and, hence, cities, not nation-states, constitute the “basic, salient entities of economic life” (Ibid: 31).

Taylor proceeds to explicitly merge Jacobs’ position on the economic centrality of cities with an historical account, borrowing heavily from World Systems theory, arguing that the history of the modern World System of capitalism has coincided with the sequential productions of a metageography, defined as an “unexamined spatial discourse that provides the framework for thinking about the world across the whole gamut of human activities and interests” (Taylor, 2004: 180).  Taylor uses the conception of metageography in order to argue that the intensification of international connections associated with globalization, and, critically, the role of cities in facilitating such connections, poses the opportunity for the transition away from a metageography configured as a mosaic of territorially bounded nation-states and national macroeconomies to one configured on interregional/interurban networks.  Notwithstanding the centrality of economic processes to the creation of such a metageography, Taylor elaborates on effects to non-economic processes, introducing, as outcomes of globalization, the potentiality of new interregional networked political sovereignties and the construction of cosmopolitan cultural self-identifications transcending, at least in part, national/ethnic identities linked to the political processes of the nation-state.  The conception of globalization advanced by such a metageography paints the image of a post-national world of networked cities in which the nation-state, as an accepted receptacle of political power and coherent macroeconomic organization and the wellspring of cultural self-identification since the time of the Treaty of Westphalia, becomes a structural anachronism, unnecessary to the workings of interregional economic networks and, perhaps, an impediment to development of networked forms of urban or interurban political sovereignty.

It suffices to say that Taylor’s post-national metageography currently manifests itself only as a theoretic argument seeking the moment within which to become an “unexamined spatial discourse” at the level of everyday human existence.  That is to say, using language heretofore advanced, Taylor’s metageography performs a reality to whose formation it can, in fact, indirectly contribute.  It represents, moreover, the performance of a reality that I mean to take seriously in advancing my own theorization of cities in relation to globalization. 

It would be a serious mistake, on my part, to write out the nation-state as a meaningful political agent in processes constituting the city.  The possibility for development of interregional economic networks, in particular, requires that no irresolvable obstructions exist to the extension of networks.  Such a requirement necessitates the presence of long distance transmission infrastructures, but it also presumes the existence of continuous expanses of Cartesian space for transmission pathways within which political processes will secure/facilitate rather than impede transmissions.  One extreme version in which long distance commercial network development achieved facilitation by means of a particular set of state political processes is evident in development of the “Silk Road” through central Asia in connection with the establishment, by military means, of narrow and tenuous political networks by divergent Mongol khanates in the Thirteenth century (Allsen, 1997).  In the contemporary context of interregional economic network development, the nation-state, as the dominant form of political organization, generally maintains the capacity to provide the space necessary to develop and stabilize over longer periods such interregional transmission pathways. 

Moreover, to the extent that contemporary interregional commercial exchange networks have been characterized by a trend of liberalization, it must be acknowledged that such liberalization has been effected, in part, through political processes by individual nation-states, engaging in bilateral and multilateral deliberation with other nation-states to permit the penetration of their boundaries by particular network forms.  The World Trade Organization (WTO), as an organizational entity governing interregional commercial networks, operates through the participation and at the behest of nation-states, which constitute its members.  Interregional network development at the present time is predicated, in part, on, rather than precluded by, the existence of nation-states.  As a practical matter of analyzing the position of the nation-state within contemporary globalization, this project, thus, concurs more strongly with the position of Hirst and Thompson (1999) in their acknowledgement of the continued political vitality of the nation-state under globalization than with Ohmae’s (1995) enthusiastic pronouncement of the nation-state’s death.                       

            On the other hand, arguing from an analogous position to that of Jacobs and Taylor, it is one thing to concede a critical role to the nation-state in facilitating interregional network development, but it is quite another thing to imply that an interregional network is any more significant if it is, simultaneously, international.  The latter modifier describes a relationship between nation-states that I do not intend to emphasize or even utilize again in defining globalization as a set of processes connecting regional spaces focused on cities.  In concluding this section, I mean to explain why I think a city-centric/interregional, but not international, conception of globalization is important, quite specifically, to Marxism.

            To the extent that Marxism, as a theoretic tradition centered on class processes, contains a commitment to internationalism, witnessed in various incarnations of international working class movements from the 1860s to the present day, it must come to terms theoretically with the nation-state.  However class is theorized in Marxist accounts to conceptualize a fundamental commonality that transcends political/jurisdictional boundaries on a global scale to unify all workers/producers of surplus value engaged in capitalist production processes, Marxism must recognize, at least as a condition of struggle for international solidarity against bourgeois nationalism, the existence of traversable national boundaries, continuous state spaces contained by such boundaries, and political and/or cultural processes instituting them.  In defending internationalism, Marxian theory, at least implicitly, draws a global geography of nation-states and, in the case of certain theorists (e.g. Luxemburg, 1976), culturally (ideologically, linguistically, etc.) autonomous nationalities contained within the political/jurisdictional boundaries and integrated domestic market space of multi-national states (e.g. autonomous Polish national/cultural space integrated economically and politically into the state space of late Tsarist Russia).[20]  Succinctly, Marxist internationalism accepts a link between, on the one hand, the political structure of the nation-state and/or the cultural structure of national/ethnic groupings and, on the other hand, the geographic identification of continuous territorial spaces organically belonging to the nation-state,[21] even as it attempts to forge transnational linkages based on class-identifications. 

            The problem with this territorialization of the nation-state is that it often co-mingles urban and non-urban spaces without adequately conceiving of the extent to which the nation-state constitutes a complex envelope of heterogeneous spatial forms, unevenly connected to economic, cultural, political, and physical processes transcending the jurisdictional boundaries of the nation-state.  Cities and their proximate hinterlands are exposed to the effects of processes transcending nation-state boundaries in qualitatively different ways than relatively disconnected non-urban spaces.  These differences both confer on agents in cities a kind of relational power over economic processes not possessed by agents in distant non-urban spaces[22] and exact from cities, as the price of openness, a degree of vulnerability to the effects of cultural processes manifesting a corrosive and subversive impact to cultural processes on which the nation-state relies to cement its hold over Cartesian state space.  To a great extent, the role of cities in interregional transmission processes implies that cities must function as gatekeepers for all regions of the nation-state relative to the effects of processes transcending nation-state boundaries. 

Cities like New York, London, and Paris accumulate linkages to cultural, political, and economic processes outside of the nation-states within which they are located whose effects will never reach Ohio, Lancashire, or l’Haut Rhône in an unmediated form.  Direct investments of capital by foreign capitalist firms to facilitate commodity production and extraction of surplus value in particular city regions transform, however imperceptibly, the degree to which cities integrate multiple national economies.  The organization of surplus value for such firms reflects a multinational flow of distributed surplus values, from production/appropriation nodes in particular city regions, to capitalist boards of directors at receiving nodes across national jurisdictional boundaries.  In this manner, the conditions of existence for surplus value production in one city region include economic (e.g. procurement of finance) and political (e.g. state monetary, fiscal, and regulatory policy) processes in other nation-states (Ruccio et al., 1991: 34).  This integration in the organization of surplus value across national jurisdictional boundaries does not, however, imply the integration of two national economies, but the integration of geographically dispersed sites/nodes of class processes in or through two city-regional economies, contained by different nation-states.  Again, flows of money capital from a Chinese capitalist firm in Shanghai invested in commodity production in the New York or Los Angeles regions do not necessarily or immediately integrate rural economies in Oklahoma or Indiana into a single globally extensive economic network connecting them to rural economies in Shanxi or Shandong province.     

In addition to flows of capital, extensions of class networks across national jurisdictional boundaries involve flows of labor power, some professional, some unskilled.  Migration of labor power into a city region can reshape practices in heterogeneous, skill-specific markets, transforming local labor-management relations (e.g. strengthening or undermining labor organization, raising or lowering skill-specific compensation rates).  In this regard, Sassen’s (1990) claim that the development of advanced producer service production processes in certain cities transforms local structures of formal and informal market demand for labor power, on the one hand, enhancing demand for skilled professionals, technicians, and unskilled consumer service workers and, on the other hand, eroding demand for semi-skilled manufacturing workers, constitutes an important argument on the effects of globalization on urban markets for labor power.  Assuming the basic validity of Sassen’s hypothesis for purposes of argument, expansion of demand for unskilled migrant labor power in informally structured markets for domestic cleaning and childcare services in Boston does not imply that similar demands for labor power should arise in rural Maine or even in relatively isolated, deindustrialized manufacturing towns in Western Massachusetts like Pittsfield.  

Beyond the economics of markets for labor power, human migration imposes relevant cultural effects on cities.  Immigrant populations, engaged in cultural processes at variance with those of host nation-states, contribute to a cumulative corrosion/subversion of the geographically defined cultural connection of the city to the nation-state.  In some cases, the persistence of such cultural practices or, rather, the appearance of persistent cultural differences between domestic and immigrant populations/failure of assimilation may reflect the presence of survival strategies by immigrant groups in negotiating insoluble racial differences relative to “native” populations.  However such differences arise, enhanced migration to cities may yield a kind of cultural disarticulation, or, more properly speaking, the articulation of city spaces into multiple, overlapping heterogeneous transnational cultural networks at least partly liberated from traditional national territorial spaces. 

At issue, as economic, political, and cultural networks extend to greater lengths/become more global, is the capacity of nation-state and national cultural processes, attaching specifically urban populations to territorial state space, to assimilate, fully or at least marginally, new, more intense waves of transnational cultural network formation, accompanying the formation of new transnational economic networks through the space of cities.  That is to say, the engagement of city economies with global networks in the migration of labor power and the introduction of transnational ethnic/cultural processes forces nation-states to synthesize divergent trajectories in domestic and multiple, heterogeneous transnational cultural processes in order to maintain the cultural-geographical suturing, within the territorial envelope of the nation-state, of cities to non-urban spaces, relatively disconnected from globalization.

The problem that Marxism, and Marxist internationalism in particular, thus, faces is to articulate an understanding of globalization that will perform the partisan reality it seeks to advance, relative to class processes.  Namely, Marxism seeks to create spaces within a continuously developing global economy for communism, as the collective appropriation and distribution of surplus labor by its producers.  The status of the nation-state in relation to this larger partisan project is theoretically ambiguous by virtue of the complex, overdetermined relationship between state political processes and class processes, a relationship irreducible to last instance determination by the economic.  Fundamentally, at stake is the capacity of the nation-state to secure particular political conditions of existence for the development of communist class structures, either by actively privileging communism by means of policy (e.g. explicit state recognition of collective surplus appropriation/distribution, preferential tax policies, etc.) or by supporting a liberal economic developmental environment that does not politically or economically stifle the development of communist structures.

It is a contention of this project, one supported by the perspectives of Jacobs (1984) and Taylor (2004), that cities, not nation-states, exist as a geographically central focus for the constellation of economic, political, cultural, and physical processes involved in globalization, and, in fact, that globalization merely describes connections that cities have, by definition, always made with other city-regional economies.  Moreover, I will commit the argument in chapter 3 to the proposition that in such an environment of free and furious economic development, induced, in part, by globalization, radical degrees of class structural heterogeneity are conceivable.  On the other hand, the creation and policing of national jurisdictional boundaries, however liberal in nature, and concomitant cultural processes, tying economically disconnected urban and non-urban populations together through a fictive connection to a common land space, hinders, at least in some degree, the development of such vigorous environments.  In this sense, concurring at least in part with Hardt and Negri’s (2004) imagery of democratic possibilities in articulation of the global multitude and the destruction of existing forms of national sovereignty, this project asserts the need for Marxian theory to sever its connection to an internationalism predicated on a meta-geography of nation-states and pursue a more rigorous, borderless globalism.    

---

[18] Approaching from the perspective of ANT, Law (2004) suggests, in a similar manner, that the global “is specific to each location, and if it is bigger or smaller then it is because it is made bigger or smaller at this site or that” (2004: 24, emphasis in original).

[19] This imagery of globalization conforms, at least in part, with Hardt and Negri’s (2001) conception of the construction of Empire through the production of impersonal network structures in which “there is no place of power – it is everywhere and nowhere” (2001: 190).  The emphasis in this project in theorizing interregional networks is, again, clearly different, however.  As with Castells’ network society (2000A) and Deleuze and Guattari’s (1987) concept of rhizomatic, fluid networking, Hardt and Negri’s theorization of Empire seeks to dissolve fixed spatialities and emphasize the placeless, non-stationary character of information flows/exercise of power by capitalist firms and undifferentiated state actors– this project seeks to articulate the nodal spatiality of the networks and argue the critical role played by cities in such articulations, especially when the networks are globally extensive.    

[20] Neither Luxemburg’s affirmation of cultural autonomy for Polish regions in pre-World War I Russia nor Lenin’s pre-1917 support for complete national self determination for Poland and other Russian ethnic minorities (Lim, 1995) appear to support the contention by Hardt and Negri (2001) that Marxist internationalism, as a matter of organizational practice, was predicated on an anti-nationalist position.   

[21] I am indifferent, in this respect, to the question, treated by Balibar (1990), concerning whether nations, as ethnically or otherwise culturally defined groupings, give rise to political states or whether state political processes, under selectively inclusive, liberal politics, give rise to nations, reflecting Anderson’s “imagined communities” (1991).  The key point that I mean to emphasize is that the cultural production of a symbolic geography of the nation-state operates as a condition of existence for the nation-state, however it originates. 

[22] What I have in mind here is the degree of power over, for example, financial flows, enjoyed by financial sectors firms in the City of London, relative to industrial or retail/wholesale firms in the deindustrializing Northern English counties (Massey, 2007).  As Gibson-Graham (2006B: 23-51, 127-163) argue, to the contrary, agents in relatively disconnected, non-urban regions are not powerless in the face of global flows of finance capital, information, other use values, and people.  Economic agents in the Latrobe Valley of Australia and the Pioneer Valley of Massachusetts in the U.S. possess a different relational capacity to engage with globally extensive economic processes in alternative ways, but, notwithstanding, I mean to probe the particular aspects of the relational power wielded by cities because I find it particularly relevant, though by no means essential, to the manner in which other regions of nation-states engage with globalization.  

Cities, Networks, and Globalization III

 Cities as Connectors

This section constitutes the beginning of my theorization of the city as a spatial form, emanating from the preceding theorization of networks.  Two introductory points need to be advanced in making the transition in this section from the ontological abstraction of networks to develop a relatively concrete account on the city.  First, reiterating the consequences of approaching the city from a framework of ontological overdetermination and, further, from an epistemological position conceived as performative, in the sense described in chapter 1, this project rejects the notion that any objective definition of the city can be articulated by theory or analysis.  I advance a definition of the city in this and the succeeding sections of this chapter with the explicit intention of articulating persuasive arguments on the city that can inform an analysis of heterogeneous class structures in cities.  In this respect, I mean to be quite explicit in conceding the irresolvable complexity of the city as a spatial form against which my theoretic intervention is strictly partial and partisan in its aims.  There can be no objective general theory of the city.   

            Second, the material in this and the next section seek to define characteristics of the city related to its connections to processes external to its boundaries.  Disaggregating this conception of a relationship between cities and the outside requires that I inquire into the notion of the city as a bounded form with a well-defined interior and exterior.  Further, in holding to the conception of the networked spatio-temporality of processes introduced in the previous section, it requires that I inquire into the form of transmission vehicles/nodes and pathways connecting relatively stationary network nodes contained by the city to those external to its boundaries.  These two tasks will be dealt with in this section.

Defining Boundaries

The theoretic conception of the city that I will advance in this project presumes a well-defined interior and exterior and, further, a defined regional geography constituting space that might be labeled a metropolitan hinterland.  Such an image, in turn, presumes the existence of boundaries, capable to establishing a coherent separation of that which is inside the city from that which is outside.  The city, in this sense, exists as a bounded spatial form, but the introduction of boundaries raises two sources of ambiguity that I must resolve.  First, what kind of boundaries am I introducing to delimit the city?  Second, insofar as the city is a theoretic articulation and theory is conceived in performative terms, why am I articulating a bounded city as opposed to a city conceived as an open, unbounded spatial form?

            In responding to the first of these ambiguities, the boundaries of the city that I have in mind are not the boundaries drawn by political processes defining the spatio-juridical limits of authority for particular political organizations (e.g. municipal or regional/metropolitan governments).  Acknowledging the validity of such a definition of boundaries and their importance to theoretic/analytic accounts attempting to come to terms with questions involving the exercise of political authority in addressing problems like regional economic development or ecological waste control, such a reduction of boundaries to those drawn by political processes will not address the particular needs of this theoretic account.  Specifically, this account seeks to answer questions concerning the spatio-temporality of class structures (i.e. the spatio-temporal dispersion of class networks).  In particular, when can I assert that a particular spatial agglomeration of class processes and inter-connected class structures is contained by some spatial form called a city?  If I assert that the conception of the city as a container, with defined boundaries, must, in some way, be determined by all of the processes it contains, then I should be able to articulate a theoretic account linking the boundedness of the city to its containment of class processes.  In what sense can the city, therefore, be said to be bounded with respect to class processes?  This is the question I must answer.

            The conception of boundedness with respect to class processes that I will advance here borrows, in part, from basic conceptions of spatial economy, concerning, in particular, agglomeration economies.  The understanding of the economic theory of agglomerations from which I am proceeding subsumes a wide range of divergent theories from Von Thünen’s (1966) abstract theorization’s on the spatial distribution of economic processes in isolated agrarian economies, to Alfred Weber’s (1929) theories of industrial location, Isard’s (1972) attempt to synthesize diverse, mostly German, strains of location theory into a general theory of location, and, most recently, Krugman’s (1991; Fujita et al., 2000) efforts to develop a Neoclassical theory of geographical economics based on monopolistically competitive pricing models.  My primary interest, however, is with central place theory, developed separately in works by Christaller (1966) and Lösch (1967), which posits, with varying degrees of specification, an account of the relationship between cities and their peripheral hinterlands.  My use of central place theory as a “a classification scheme, a way of organizing our perceptions” (Fujita et al., 2000:27) of spatial economic organization runs counter to the purposes embodied in most recent Neoclassical strands of geographical economics, but, generally, will not hinder the manner that I will incorporate it, as an implicit background, into an explanation of the city.

            My intention, in borrowing the basic idea of agglomeration economies, is not to dabble in the sorts of abstract Neoclassical assumptions, mathematical general equilibrium-oriented models of perfect or monopolistic competition, and stochastic perturbations necessary to distribute homogeneous firms unevenly across a flat plain with homogeneous characteristics.  I, likewise, do not intend to engage in micro-founded analyses to determine how individual agents, displaying characteristic features of rational choice, produce agglomerative patterns.  Rather, I intend, simply, to argue that, however such distributions arise, a dense grouping of economic processes, with local connections to a range of other non-economic processes, in a particular, arbitrarily compact space constitutes a spatial form that I will label a city.     

In its simplest terms, a city, within this framework, constitutes a basin of attraction for economic processes.  Such basins of attraction may arise, in part, from the effects of economies of scale or scope in the regional agglomeration of agents in economic production (e.g. firms), reducing average production costs per unit of output.  One source of scale/scope economies may arise from minimization of transportation expenditures between sites of production in integrated relations of supply/consumption between producers.  Agglomeration of production agents on this basis, further, generates a centripetal pull of consuming agents (e.g. households) toward the basin of attraction, reinforcing the advantages of central location in the basin by making it a center for (market or non-market) distribution and consumption.  This centripetal pull generates relatively central and peripheral spaces of economic processes.

On the other hand, borrowing from a well developed Marxist literature on the spatial effects from imposition of differential rent (Marx, 1991: 779-811[1]; Harvey, 1973: 176-194), the centripetal pull of a basin of attraction must be unevenly balanced, relative to individual producing and consuming economic agents, by a centrifugal push.  Rent arises, in this circumstance, because external economies/reduced average production and distribution costs generate super profits, relative to socially average/normal profit rates, for agents of production located in or near the basin of attraction and the capacity for private landholding and leasing enables landholders to extract these super profits as a condition of providing advantageous locations.  Perfect extraction of super profits as differential rent constitutes an unnecessary assumption in this circumstance.  It suffices to argue that any imposition of rents on producing or consuming (e.g. on favorably located residential space) will generate some degree of centrifugal pressure. 

In this regard, a kind of spatial dialectic must operate, in which the possibility for economic advantages from location close to the basin of attraction (i.e. external economies of agglomeration in minimization of distance to distribution/market exchange sites in the basin of attraction) are unevenly counterbalanced by the imposition of rents and the consequent extraction of super profits from producing and consuming economic agents.  This interplay of centripetal and centrifugal pressures around basins of attraction for economic processes precedes from a general assumption, evident in both Neoclassical and Marxian approaches to urban development, that competitive, cost-minimizing factors (e.g. relative to transportation costs) should drive the location decisions of firms and households.[2]  The image of space generated by this kind of theorizing appears as a field of unevenly distributed economic agents, in which some corners of space exhibit dense concentrations of firms (and activities, like market exchange, associated with the presence of firms) and consuming agents, and other areas appear relatively empty.[3] 

If I begin from this initial, abstract image of agglomeration economies, in which, to apply terminology consistent with the Marxian class analytical lexicon, some relatively dense collection of class processes, producing, appropriating, distributing, and receiving quantities of surplus labor, occupies the relatively compact space of a basin of attraction, then I can, further, generate an abstract image of an outside, of spaces that are not contained within such an agglomeration.  Strictly speaking, such an image must consist of two general types of spaces: spaces in which the density of economic processes across the same spatial area is relatively lower and spaces in which the density of economic processes across the same spatial area is equal or relatively higher.  At this level of abstraction, the former types of spaces might be labeled non-urban and the latter types might constitute other cities. 

Proceeding from the assertion that what separates a city from other cities are intervening fields of non-urban space, cities, in this perspective, must be bounded by some form of gradient across which the density of economic processes distributed for a given area declines.  Simultaneously, such a gradient must constitute a register of average differences in the imposition of differential rent, such that rental rates, on average, should be expected to decline directly with distance from the basin of attraction.  Such gradients exhibit arbitrarily steep declines in the density of economic processes and rental rates in which the steepness may be continuous or discontinuous.  If the separation between a city and its outside is constituted by a relatively steady and continuous or a relatively abrupt and disjunctive decline in the density of economic processes separating the center from its periphery, then the notion of a boundary must become as amorphous or as well defined as the gradient against which densities decline.  The visual analogy to be drawn in conveying the mapped image of the city as a relatively dense economic center with an amorphous downward sloping gradient in all directions is that of a topographic surface with contour lines designating changes in altitude (e.g. from the summit of a mountain to its base).  Such surface mappings may represent long continuous slopes or the presence of an abrupt precipice.     

This image of boundedness may, further, be complicated if economic processes are concentrated in multiple, closely grouped localities separated by intervening spaces with relatively lower densities of economic processes, representing a multi-polar/multi-centered city whose character as an integrated spatial form would be constituted by the interconnection of its multiple basins of attraction for economic processes.  Such cities would contain cleavages characterized by lower densities of economic processes, bordering basins of attraction and opening out into spaces of successively lower densities in geographic transition to the metropolitan periphery.  Thus, the larger geography of a multi-polar city constitutes a discontinuous surface of mixed, alternating densities of economic processes bounded by a single, all encompassing gradient of indeterminate slope/steepness.  In this sense, the boundedness of cities, as economic agglomerations, by gradients in relation to spaces with lower densities of economic processes applies both to simple agglomerations with a single basin of attraction and to integrated forms with multiple, interspersed basins of attraction. 

This introductory conception of the city as an agglomeration of economic processes, and, in particular, class processes, does not present an obvious connection to the theory of networks advanced in the previous section.  More pointedly, it suggests that the city exists as an economic interior separated from its outside, if not by walls or jurisdictional boundaries than by a declining density of economic processes.  This is not the condition in which I intend to leave the problem of defining boundaries.  On the contrary, I want to reframe the city, as an agglomeration of economic processes, in order to propose its relatedness to the outside, suggestive of permeable boundaries.  At this point in my argument, I need to reintroduce the ontology and spatio-temporality of structures and inquire into how networked processes transcend individual cities.

It may be the case that a particular economic organization defined by a class structure (i.e. a firm) has all of the sites of its class processes (i.e. its surplus production, appropriation/distribution, and receiving nodes) located in the same agglomeration, alongside the sites of numerous other class processes belonging to other class structures.  It is also, however, possible that the class processes in a given class structured organization will have spatially dispersed sites, with certain surplus production processes occurring in one agglomeration, other production processes occurring at sites in the non-urban periphery/hinterland of the agglomeration, the organization’s headquarters facility (a site for receiving and distributing surplus) in an entirely different agglomeration separated from the production sites by hundreds of miles, and recipients of surplus labor, producing a range of conditions of existence for the surplus production process, dispersed across dozens of different sites in many different agglomerations and non-urban locations.     

If the organization of surplus labor is networked, in this manner, across multiple agglomerations, then there can be no question that the economic processes constituting the city as an agglomeration exist in continuous interaction with the city’s outside.  The geography of the city is a relational geography.  Every economic process contained by an agglomeration exists in a structural and spatio-temporal relationship to other processes within the agglomeration and to processes outside of it, in peripheral non-urban space/hinterlands and in other cities.  The gradient that constitutes a city’s boundary, relative to economic processes, must, therefore, be a permeable boundary traversed by network pathways open to the flow of people, physical use values, and information by a variety of relatively mobile transmission vehicles.  The permeable nature of the city’s boundary gradient and the capacity of networks to transcend the gradient implies, further, that the city may exist as both an agglomeration of economic processes and as the spatial milieu through which those processes become connected to other processes. 

            The definition of boundaries that I am, thus, advancing in this section is represented by a gradient, along which the density of economic processes occurring across a particular geography declines with distance from agglomerative centers/basins of attraction of economic processes, identified with cities.  Such a gradient may exhibit a smooth or a disjunctive decline.  Further, cities may contain a single basin of attraction or multiple basins of attraction interspersed by internal gradients, constituting multi-polar cities.  The economic processes, including class processes in the organization of surplus labor, contained by cities, lastly, do not exist in isolation, but manifest network connections that will generally extend outside of the city.  Thus, the city, itself, must be understood, in part, as an agglomeration of economic processes organized into structures of processes and, spatio-temporally, networks.  In view of the relational character of the networked economic processes, the city is constituted not only by the agglomeration of economic processes internal to its boundary gradient, but also by the external processes to which these internal economic processes are connected across the permeable space of the boundary gradient. 

            The remaining question that I must answer before moving forward concerns why I want to theorize the city as a bounded form, however permeable to penetration of its boundaries and determination from its outside.  The simple answer to this question is that I want to apply some conception of place through which I can differentiate cities from their non-urban hinterlands and from other cities.  Elaborating, I am trying to steer a middle course between two possible alternative conceptions of place in which, on the one hand, the internal development of place overwhelms its connectedness to its outside, and, on the other hand, the entire existence of place is constituted by its outside.  In the former approach, the boundaries of the city entirely inhibit its capacity to be continuously transformed by processes in other places.  In the latter approach, the boundaries of the city become non-existent and the city, itself, becomes a pure, placeless conduit for external flows of people, of commodities, and of information.  In relation to both extremes, the approach to boundedness that I seek to advance might be described as a relational conception of place, along the same lines as Massey’s understanding of a “global sense of place” (1994: 146-156).  In articulating this idea, Massey argues at length that:

(t)he globalization of social relations is yet another source of (the reproduction of) geographical uneven development, and thus of the uniqueness of place.  There is the specificity of place that derives from the fact that each place is the focus of a distinct mixture of wider and more local social relations.  There is the fact that this very mixture together in one place may produce effects which would not have happened otherwise.  And finally, all these relations interact with and take a further element of specificity from the accumulated history of a place, with that history itself imagined as the product of layer upon layer of different sets of linkages, both local and to the wider world (1994: 156).

 

In the terms that I mean to describe the city and its boundaries, it would make little sense to argue that the identity of a place is constituted as a “mixture of wider and more local social relations” in the absence of some means to differentiate a geographic locality from its outside, however amorphous the boundary is between the local and the non-local.  More generally, a relational conception of place must, in some way, comprehend the construction of a place through its myriad economic and non-economic processual relationships to its outside without simultaneously dissolving it into a geography of flows between places.  That is to say, there must be an inside of processes with which the outside relates.  My emphasis on networked processes seeks to address this need to define the city as a place constructed in relation to other places but with a definite internal existence of its own. 

Transitional Nodes and Short and Long Range Connections

If the processes contained by the economic agglomeration that I will label the city are themselves contained by structures that constitute spatio-temporal networks, then the question further arises concerning how relatively stationary network nodes in the city connect with other relatively stationary nodes outside of its boundary gradient.  At an abstract level, network theory, in the tradition of ANT, requires that there must be some transmission vehicle that is, itself, a node in the network.  In particular, Latour (1987: 227) interjects the conception of immutable mobiles, describing objects in transmission in Euclidean space (or, across historical temporality[4]) that retain their configurations as actors within a network.  The underlying principle for this conception is that the space of an actor network is not Euclidean/Cartesian[5] space – actor network space is a relational space constituted by the hybrid connections of actors. 

            Using a popular metaphor introduced by Law (1986), a Fifteenth century Portuguese ship constitutes a network of human and non-human actors that, to the extent that none of its pieces are removed or their particular mobilizations fail to achieve a desired end, retains its shape while navigating the Cartesian space from Lisbon to Calicut (Law and Mol, 2001: 611-612).  The ship is, therefore, mobile in Cartesian space but immutable/immobile with respect to the configuration of actors constituted both by the ship’s internal parts and all other actors facilitating the processes by which its transmission through Cartesian space occurs (e.g. the financiers, merchants on both terminal ends, navigational knowledge embodied in practical experience and learning, favorable climate conditions, absence of intervening hostile attack, etc…).  The actor network maintains an existence that can be regarded, in terms of Cartesian mappings of space, as placeless.  Rather, the spatiality of actor networks is topological, implying that the technologies configuring transmission vehicles fold and otherwise displace any universal set of measurement between points in Cartesian space, enabling objects in transmission to hold their integrity as objects, on the one hand, and, on the other hand, transform the temporal distance connecting points in Cartesian space (i.e. “to annihilate (Cartesian) space by time,” to compress space-time) (Marx, 1993: 524-525; Harvey, 1989A). 

            At the outset, I want to acknowledge that there is something to this understanding of the relational properties of network space and the necessity of reconsidering the intercourse between the spatio-temporality of networks and the spatiality of geographical mappings.  If I am going to continue to take seriously the basic ontological notion that space and time are the dimensions of processes, then no geographical mapping, asserting objective and universal validity, can trump a relational analysis on the connection between processes separated by distance and time and connected, as mutually constitutive processes, by a transmission process.  In this sense, the characteristics of the transmission process and, in particular, its capacity to achieve topological transformations of space-time to produce a different network spatio-temporality is critical.

            Interpreting these insights with respect to my developing definition of the city, network nodes contained by the city and connected to points outside of the city via transmission processes must exist in heterogeneous, complex relationships with Cartesian mapped space.  Every kind of different transmission process bends space-time in a particular, unique way to facilitate the transmission of objects and effects from every networked process contained by the city.  Consequently, the city exists as a complex intersection of all these transmission processes with their particular, broadly heterogeneous relational network mappings. 

            This image of the city as an area in Cartesian space where space-time becomes bent in myriad different ways by connection to the outside represents an alternative relationship to Cartesian space from that implied by my consideration of the boundary gradient.  The gradient represents a diminution of the concentration of economic processes along an interval from a given basin of attraction, reflected in a certain expanse of Cartesian space.  Its mapping is relational but it maps the capacity of the city to be distinct from other places on its outside.  The relational mappings of networks with transmission processes emanating from the city penetrate the boundary gradient as if it was not there precisely because they manifest a different kind of intercourse with Cartesian space from that of the gradient.  I am, thus, dealing presently with three distinct kinds of relational space in describing the city and speculating on the relationships between each: the space of geographical surface mappings (Cartesian space)[6]; the topological space-time of networks; and the variable quantitatively imprinted surface mapping of basins of attraction and gradients relative to the density of economic processes.  Cities exist simultaneously in all three forms of relational space, and the intercourse between these forms of relational space constitutes the particular position of the city in relation to its outside. 

            At present, focusing in particular on the topological space-time of networks, I want to argue that there are some characteristics of the transmission processes emanating from the city that makes cities different from non-urban space, particularly those of their hinterlands.  Specifically, certain transmission processes emanating from the city must transform distances in Cartesian space in a relatively more radical fashion than most transmission processes emanating from non-urban spaces.  That is to say, network connections emanating from cities must be able to extend over longer distances and/or accelerate the movement of objects or effects over longer distances relative to transmission processes emanating from non-urban space.  The argument is simply that space-time compression, as the capacity to reduce the distance between two areas in Cartesian space by accelerating the rate of movement along the pathway, is a process primarily associated with cities.  I mean to take it a step further, however, and to argue that space-time compression is a definitive characteristic of the city – that a city is not a city if transmission processes emanating from it do not compress space-time in radical ways.  This argument relies, in particular, on the idea that the transmission processes connecting stationary nodes in cities to stationary nodes in their hinterlands take a particular form that is, in relative terms, different from the transmission processes connecting two different cities.

            In considering here the transmissions of networked processes, I am speaking, in particular, of the transportation of material objects/physical use values and people and about the communication of information over relatively long distances.  Such transmissions are certainly not exhaustive of the effects generated by material processes, but they are a highly evident starting point for discussing space-time compression.  Reasons exist why particular processes involved in long distance transportation and communications might be regionally concentrated in cities.

             First, evaluated both within from Marxian and Neoclassical spatial economic perspectives and from the perspective of network theory, the spatial displacement of use values, people, and information is costly.  Resources must be expended to move material objects from one point in Cartesian space to another.  From a spatial economic perspective, Isard (1972: 79) expresses this costliness in reference to transport inputs, denoting the quantity of resources necessary to move a unit weight over a unit distance.  No such concept is explicitly developed in Marx’s (1992: 225-229) consideration of the costs of transportation, but his analysis demonstrates the same concerns with “cubic content and weight” in addition to the distance of movements.  A governing component in both these considerations involves the technologies employed in the transportation process and their role in determining the productivity of the transportation industry, expressed in its capacity to displace unit distances in Cartesian space at progressively accelerated rates. 

            Proceeding with a disaggregation of transportation costs from Marxian value framework, the transportation process combines elements of fixed (e.g. transportation vehicles) and circulating (e.g. fuel) constant capital and variable capital (i.e. labor power) in order to directly facilitate the transmission of materials in Cartesian space.  On the other hand, another component exists to be accounted for, in reference to the particular technologies of transportation.  Specifically, transportation and communication processes make use of specific, dedicated pathways, the signature of which is left by investment in infrastructure (e.g. rail systems, telephonic and electrical grids, water and waste movement systems, etc.).  Such infrastructures, as fixed constant capital components, must be produced and, periodically, reproduced in order to successfully execute transportation and communication processes.  Sometimes firms engaged in transportation and communications processes undertake these investments in infrastructure, while others are collectivized through states.  Infrastructures, moreover, tend to be embedded in Cartesian space.  Once constructed, a road system or an electrical grid holds to a fixed geography, against which outlying spurs can be laid but larger territorial displacements of the underlying backbone are impossible.  In this sense, the fixed Cartesian geography of transportation/communications infrastructures is critically important.[7] 

            Disaggregating transportation and communications processes, again, in reference to network theory, the same above distinction between movement and fixed infrastructure remains relevant.  First, transmissions, as mobile processes, involve continuous movements, displacing certain distances between places in Cartesian space at a particular, relatively constant temporal rate.  That is to say, a railcar may move at an average speed of 55 miles per hour on a continuous rail line from Chicago to Detroit.  Such a rate is facilitated by a particular transportation technology, the dedicated infrastructural pathways on which technologies are employed, and a range of other physical and social processes that overdetermine the vehicle’s rate of movement along the pathways. 

            On the other hand, transmission processes may involve moments of transition between particular transportation/communications vehicles, facilitating movement along distinct, dedicated transmission pathways.  An example of this transition process might involve removal of freight from a railcar to an on-road vehicle, reflecting different potential average rates of movement between places along different infrastructure pathways.  Such transitions take place at critical points in infrastructure systems that I will alternately label transitional nodes or access points (e.g. rail yards, container ports, air hubs, satellite communications relays, entry/exit points for freeway systems) that may be more costly to produce than components in infrastructure systems that facilitate continuous movement (e.g. stretches of road, rail, or fiber optic cables).  Transmission processes, particularly those involving transport and communications, may, thus, integrate distinct relatively stationary (transitional) and relatively mobile (continuous movement) nodes.  The integration of both sets of processes is required for the successful execution of a transmission across Cartesian space because the inability to utilize a transitional node/access point in a privileged/exclusive infrastructure pathway implies exclusion from mobile transmission processes along the pathway.  In this sense, the location of access points, as fixed/embedded infrastructural components, together with the relative costliness of developing such access points, regulates, to an important extent, the capacity of network transmission processes to bend space-time.

            Again, this understanding of network concepts necessarily implies that large expanses of Cartesian space, even in close proximity to the basin of attraction, may be excluded from access to transmission infrastructures by their inability to utilize access points.  For example, freeways may facilitate more rapid road based transportation.  They may, likewise, saturate the geographic landscape of cities and rudely slice through the bucolic scenery of rural pastures, imprinting on quiet countrysides the image of industrial society.  But, notwithstanding the incidental inconveniences attendant to streams of automobile traffic, areas bisected by freeways that lack entry and exit points cannot make use of such infrastructures.  At all scales of transmission (intra-regional and interregional), the configuration of diverse infrastructure pathways into “hubs,” “spokes,” and “arteries” with limited access points generates tunnel effects (Graham and Marvin, 2008: 201), through which the experience of space-time compression in everyday life varies widely based on the access or exclusion of individuals and communities from proximate transmission pathways.  Existence and proximity to a pathway capable of bending space-time neither implies the existence of the technological means to traverse it or, even the capacity to enter circulation.   

            Following the arguments on a spatial dialectic of centripetal and centrifugal pressures from the previous sub-section, the location of dedicated infrastructural pathways and access points in a given city region must both constitute, in part, the balance of agglomerative economies and differential rents, relative to the basin of attraction, and be constituted, in part, by this balance of attractive and repulsive forces.  Space in close proximity to intra-regional transportation/communications corridors and, in particular, access points must command higher quantities of differential rent because the capacity to access transmission media that will undermine the friction of distance across Cartesian space confers the potential for super profit in intra-regional exchange.  On the other hand, private or state investors in transmission infrastructures must negotiate pre-existing landscapes of agglomeration and rental rates in the production of new infrastructure and the replacement of existing infrastructure.  These issues combine with larger questions concerning the capacity of specific transmission technologies to bend space-time and the compatibility of multiple technological vehicles on a given, dedicated infrastructure pathway (e.g. the capacity of a rail system to accept new, higher speed vehicles without making substantial changes to existing tracks).

            The discussion of transmission processes, to this point, has largely concerned intra-regional connections in the relationship of a city to its regional hinterlands, alluding to the capacity of transport and communications infrastructures to allow economic processes to escape from central basins of attraction because intra-regional transport and communications technologies allow, particularly, commodity producing firms in the hinterlands to bend space-time in transporting commodities to city-based sites of market exchange.  In this manner, cities, as economic basins of attraction, must constitute the terminal points for diverse, radial transmission pathways emanating from points in their hinterlands.  Further, transmission pathways in close proximity to a city may be configured for lateral transmissions, connecting separate radial pathways to form “beltway” infrastructures of varying densities.  In some cases, these infrastructures may carve small, uneven, but clearly evident, radial corridors through the regional landscape, with highly concentrated access points in close proximity to the city and profuse lateral connections (e.g. freeway road systems, rail systems).  In other cases, the profusion of radial and lateral interconnections, especially in proximity to the city, is so dense, that infrastructures appear to completely saturate the landscape (e.g. landline telephone, fiber optic cable, and water transportation systems), with profuse, but by no means universally or evenly dispersed, access points.  In either case, the city constitutes a space in which transmission pathways and access points, using all available technologies for intra-regional transmissions, get pulled together. 

            If a city is the place where all of the network pathways within a particular region get pulled together, then it may also be the place where intra-regional pathways link to transitional nodes/access points for interregional transmission pathways over longer distances.  Whether or not the city undertakes this role depends, in part, on the costliness of transmission infrastructures and, in particular, of access points.  If infrastructure access points for long distance transmissions, like container ports for long distance freight transportation or satellite communications relay infrastructures, are especially costly to construct, then it may be impractical to construct multiple such access points within a region.  Given this condition, it is conceivable that such access points could be constructed anywhere within the region.  However, if interregional access points were located in close proximity to the city, in a location where they intersected radial and lateral intra-regional transmission pathways, then the centrality of interregional access points would minimize the costs to complete transmissions to economic process nodes throughout the region.  If the cost and locational conditions alluded to here obtain, then, limiting the terms of this argument to consideration of transmission infrastructure access points/transitional nodes, interregional transmission capacity would be largely concentrated within or in close proximity to cities.

            On the other hand, the placement of access points/transitional nodes constitutes an important process in reshaping a regional geography and, in particular, in shaping a region’s interaction with other regions.  Any space where access points to interregional transmission pathways are placed may develop into a basin of attraction for economic processes oriented toward interregional distribution/exchange.  Such possibilities stimulate local demands for infrastructure investments in the construction or renovation of roadways, air facilities, and other transportation/communications infrastructures.  Whether the placements of infrastructure succeed in stimulating aggregation economies, opportunities for super profit, and extraction of differential rent depends, however, on a long list of other economic, political, cultural, and physical processes.  Jacobs’ (1984: 105-123) criticisms of the large-scale infrastructure investments in regions lacking economically dynamic cities (e.g. the Tennessee Valley Authority), in particular, demonstrate why the placement of transmission infrastructures outside of cities may not produce viable basins of attraction.  Consequently, the argument advanced in this section, oriented strongly toward the interregional connectivity of cities as one of their defining characteristics, cannot be abstracted from the arguments to follow on the internal economic dynamics of cities as if the externally connective characteristics of cities solely determined their existence.                    

            Connecting the relatively abstract considerations of topological network space-time above to relatively concrete considerations of transportation/communications infrastructures in regions with cities, the existence of a city must imply the capacity for connections to local points and to other regions by means of technologies that will reshape the way space-time is manifest in transmission processes.  This image of space-time compression through the city accepts, in part, Virilio’s (Virilio and Lotringer, 1997: 66) contention that the city is a “box full of speeds, a kind of gearshift” such that a property of city space and its particular infrastructural architectures (e.g. satellite relays for instantaneous global wireless communication) is the annihilation of space through transmission technologies, appropriate to radically accelerated movements.  As a matter of definition, cities are areas in space where networks from disparate Cartesian geographies get pulled together through local/intra-regional and long distance/interregional transmission technologies.  The nature of these technologies changes continuously, but, provided appropriate conditions exist making it prohibitively costly for access points to pathways for radical space-time compressing technologies to be dispersed across non-urban hinterlands, cities must always be the spaces in which these access points will be concentrated.   

            Summarizing the points that I have sought to argue in this section, the city, as an agglomerative basin of attraction for economic processes, is delimited from its surrounding non-urban hinterland by a gradient along which the density of economic processes declines.  Such a gradient may, to some extent, be further represented in reference to differences in rental rates for the use of land, implying that rental rates must decline with distance from city centers.  Into this image of a city with static agglomerations of economic processes, I have sought to interject the network theory developed in the previous section to argue that processes contained within the city’s boundary/rental gradient connect to processes in non-urban hinterland areas and to processes in other cities.  The former such linkages configure a city’s larger region, inclusive of its hinterlands, as a mesh of network pathways, along which mobile transmission processes enable flows of use values, people, and information. 

            The latter linkages from the city, configured over longer distances, emanate from access points to longer distance transmission media that radically bend Cartesian space, relative to local/intra-regional network pathway media, consistent, in this sense, with theoretic conceptions of space-time compression.  Cities, thus, exist, in a static sense, as amorphously bounded areas in Cartesian space, but they also exist, dynamically, as areas in network topological space-time, within which they contain access points to transmission pathways that actively bend Cartesian space in myriad different ways, reflecting the diversity of network connections continuously forming, breaking down, and dissolving between and through cities.  The degree of access in cities to longer distance transmission media constitutes a defining characteristic of cities relative to non-urban space.  For the topological space-time of networks, non-urban spaces, including those of hinterlands, exist as Cartesian spaces to be folded over in developing connections between cities.  Wherever a network begins, if it is to extend relatively long distances (i.e. outside of a given city-regional space), it must pass through a city.  Utilizing this understanding of the role of cities as connectors in networks that extend relatively long distances via transmission processes that bend Cartesian space, I will attempt to articulate a working conception of what will be meant by globalization.

---

[11] In particular, type 1 differential rent, emanating, among other possible sources of extractable super profit, from the location of land in relation to distribution/market exchange sites and access points (discussed below) to privileged pathways for means of transport/communication. 

[12]The point, in this regard, is that the location decisions of firms and households exist in a mutually constitutive relationship – the centripetal pressure on firms, relative to basins of attraction, only exists because there is a centripetal pressure on households seeking opportunities to consume use values and, if necessary, to alienate their labor power through market exchange.  The relationship between producing and consuming economic agents constitutes a basin of attraction as a site of distribution/exchange, capable of exerting centripetal pull and, therefore, potentiating the extraction of differential rents in inverse proportion to the distance from the basin as a counteracting centrifugal push.   

[13] This imagery roughly conforms to the central place theory advanced by Lösch (1967) and modified in Isard’s (1972: 269-281) depiction of a Löschean system under less restrictive abstract assumptions.  It is unnecessary and somewhat counterproductive to my project to impose the geometric construction of optimally scaled hexagonal market areas, characteristic of a Löschean system – my intention is not to theorize an economic geography structured by rigidly bounded spaces, however liberally defined in scale, but to theorize the interconnection of networked economic processes across largely permeable boundaries.  

[14] The particular context of Latour’s (1987: 226-227) interjection of the immutable mobile concept concerns the assemblage of astronomical data, contemporaneous to the Sixteenth century Danish astronomer Brahe, and Copernican and Ptolemaic theoretic materials in written forms.  These sources constitute the assemblage of a network extending across both time (from at least Ptolemy’s time) and space (the contemporaneous space of continental Europe), in forms capable of being mobilized by Brahe to produce accurate planetary mappings.

[15] In using this terminology, I am alluding to mapped geographic space that can be subjected to a fixed system of Cartesian coordinate (XYZ-dimensional) mapping in which points or fields can be measured or connected in accordance with the rules of Euclidean geometry (spaces that can be treated as if they were flat, non-spherical surfaces).  Henceforth, I will simply regard this as Cartesian space, in part, because the term appears to generalize more, allowing for the use of spherical/non-Euclidean geometric descriptions. 

[16] In this regard, Cartesian space must be regarded as a kind of relational space relative to some representation of space through a coordinate system.  Notwithstanding any claims to the contrary, such representations are not objective, but theoretically constructed, partial, and partisan.  Thus, I will label Cartesian space as a kind of relational space. 

[17] Graham and Marvin (2008: 190-202) provide a good summary of political economy perspectives on the relevance of fixed, territorially embedded transmission infrastructures.