Cities, Networks, and Globalization II

Theorizing Space: Sites, Networks, and Nodes (Conclusion)

Situating Process Networks against Actor Network Theory (ANT)

The network theory advanced above seeks to remedy a perceived problem introduced by theorizing the city, as a spatial form, from an ontologically overdeterminist framework.  Specifically, if all processes exist ontologically in mutually constitutive/mutually constituted relationships and space-time exists as the dimensionality of processes, then overdetermination must presuppose an explosive transmission of effects from all processes across a seamless universal space-time, as the spatio-temporal dimensionality of all processes.  Such a universal space-time cannot be meaningfully theorized without abstracting processes and their structural and spatio-temporal relationships from the structured totality and universal space-time within which they reside.  Articulation of processual networks provides a mechanism for undertaking this theoretic abstraction.  Specifically, it articulates the spatio-temporal relationships of processes in a piecewise manner, identifying relatively stationary nodes and the relatively mobile nodes, connecting them and transmitting effects between the stationary nodes. 

            The general methodology followed above in describing the network concept has consciously attempted to define networks of processes through analogy to networks of actors and agencies, succinctly advanced by theorists in the developing corpus of Actor Network Theory (ANT).[7]  My rationale for developing such an analogous theory of networks is multifold, and an elaboration of this rationale demands a brief outline of ANT and an account of its commonalities and differences with the ontologically overdeterminist framework from which my account proceeds. 

            In a general sense, conveyed most succinctly by Latour (1993: 138-145), ANT theorizes material existence as an assemblage of actors, organized into hybrid combinations with other types of actors (hybrid actants).  These actor hybrids undertake acts of agency (i.e. transformations of material existence) under the influence/effect of multifarious acts of agency undertaken by many other actor hybrids at different times and in different places.  In this manner, the actor to ANT exists as an entity that is made to act by outside influences (Latour, 2007: 54-55).  On the other hand, the actions undertaken by an actor do not constitute unmediated outside imperatives – actors maintain the freedom to mediate or translate the imperatives that they internalize and to act in accordance with these mediations (Law, 2003: 5-7).  Agency, within this framework, always constitutes a response to internalized imperatives, where actors manifest some capacity to regulate the imperatives that they internalize and the internalization of influences is formative to the individualization/subjectivation of an actor (Latour, 2007: 207-213).  The actor, as an entity, exists as a locus for constitution by outside effects, but such effects are finite and their internalization or rejection by the actor is a free act of mediation.   

Some actors impact the actions of other actors over relatively short distances while others effect actions over relatively longer distances.  In either case, the capacity to exert these effects requires a mechanism through which effects can be transmitted across space and/or time to exert their influence on other actors.  Critically, ANT theorists insist on the empirical traceability of agency between actors (Latour, 2007: 193-194).  ANT theorists describe the traceable spatio-temporal connections between actors, through their acts of agency, by use of the network concept.  That is to say, networks represent an ontological tool to describe an empirically traceable material relationship between actor hybrids, in whatever form such a relationship takes.  Such networks only manifest durability, moreover, to the extent that the actor hybrids whose agency traces them are capable of reproducing the connection over time.  The production of network connections, especially over relatively long distances, carries non-trivial costs for actors and they must be repaid for every connection to remain open over time.  Actor networks may, thus, be tenuous and fragile constructs, liable to collapse if any of the pieces assembled to facilitate a connection across space-time is omitted or otherwise breaks down. 

            This basic characterization of ANT needs to be qualified against certain other features of its theoretic development.  Notably, the actors that constitute the basic elements of ANT are not strictly human and the agency described in the construction of networks is not strictly human social agency.  Rather, ANT pursues a dissolution of the boundary between human and non-human actors, insisting on the necessity of human/non-human hybrids (e.g. humans plus machinery; humans plus theoretically produced ideas; humans plus natural materials) in the assemblage of networks capable of diffusing the effects of agency over space-time (Murdoch, 1997).  In this manner, the social processes whose effects are traceable as networks are not objects of human agency, and still less of human intentionality, but outcomes arising from the hybridization of human actors with non-human actors.  This hybridization does not reflect a pure appropriation of tools to perform ends defined by a human actor, but a mutual translation of the collective agency of multiple actors, where each human and non-human agent possesses some degree of freedom to mediate the sources of agency against which its actions are a response.[8] 

The ability of a human theorist to influence others, for example, may require the theorist to enlist and transform ideas developed by other theorists and codified in printed or electronic materials, to utilize technologies in order codify her own ideas, to access computer networks for dissemination or to rely on the actions of many other human agents utilizing other technologies to disseminate her theories.  Once codified, the theories developed by the theorist will influence individuals only to the extent that such individuals can access, read, and otherwise appropriate the ideas expressed in these theories.  They will be effective theories (in a performative sense) only if they motivate actions on the part of those who expose themselves to the theories. 

At each stage in this process, the elements the theorist utilizes (e.g. other theories in print, technologies for codification and dissemination, other human agents using other technologies and knowledges) all represent actors in their own right or human/non-human actor hybrids.  The ability of the theorist to influence others with her ideas relies on the mobilization of every actor in this actor network, where the failure of any element to be successfully mobilized (e.g. the failure of a computer to successfully transmit theoretic materials, the negligence of a human actor in handling materials) will transform/undermine the agency exercised by the theorist.  Collectively, the agency exercised by the theorist, in association with diverse other human and non-human actors, traces the network.   

The imagery of the network, thus, expressed by ANT seeks to emphasize a decentered ontology, within which, on the one hand, human actors must successfully mobilize other human actors and non-human actors in order to act and, on the other hand, the motivations for action by the human actor (or, for that matter, non-human actors) arise from myriad other sources of agency, the internalization of which constitutes the imperative for action.  Agency, in general, ceases to be coextensive with human agency, and neither human nor non-human agencies reflect pure manifestations of intentionality.

Finally, the material existence that emerges from ANT supports a kind of constructivist ontology, through which human actors mobilize non-human actors into hybrid forms to create the elements of social life and to continuously transform and, thus, reproduce nature.  Human actors produce society, but only through the mediation of mobilized non-human actors (technologies, ideas, “natural” substances).  In the latter regard, the mediations of specific non-human actors (e.g. computer systems, ideas about the intervention of animal spirits in day-to-day life, etc.) shape the existence of specific forms of society, while simultaneously shaping the forms of nature to which human actors respond.  Material existence becomes a combinatory construction of society and nature in such a way that any polar distinctions between nature and society break down, leaving a totality of actors/substances of varying complexity, free to be mobilized as mediators in collective agency.                

The most obvious point of divergence between ANT and the Althusserian overdeterminist ontology on which my theorization of networks is based concerns the role of agents/actors relative to structural processes.  The processual characteristics of material existence (i.e. the continuous transformation of substances and the diffusive effects of such transformations on other processual transformations) in overdetermination prioritize the transformation of substances/agents rather than the momentary existence of substances in forms that can be made to act.  In this manner, overdetermination proceeds from a different basic unit in conceptualizing material existence – an articulation of connections between processes is not an articulation of connections between substances/actors whose agencies may, however, constitute such processual transformations.  In the most basic terms possible, the Althusserian understanding of overdetermination and ANT proceed from different moments or modes of material existence (static substantiality versus movement and transformation).[9] 

On the other hand, the significance of this basic difference in the mode of theorization appears less clear if certain points of analogy are brought to bear.  Most notably, the basic units of overdeterminist ontology (individual processes) and ANT (individual actors) exist as loci against which multifarious outside elements facilitate transformation or, at least, its potentiality.  For ANT, the agency of an individual actor is motivated/incited by the agency of multifarious other actors, separated across space-time and connected by vehicles through which the effects of agency can be transported.  For overdetermination, the transformation produced by a given process is effected by the transformations of multifarious other processes.  In accordance with the distinctions I have made between the fields of structured totality and theoretical articulation, this means one of two things: that every process exists as a locus for overdetermination by all other processes in structured totality; and that theorized connections can be made between processes, separated across space-time and connected by transmission vehicles, in which possibly uneven mutually constitutive relations can be specified.  Leaving aside the ontological background of structured totality for the moment, an analogy can be made regarding the complex determination/incitement of theorized units.

Noting the analogy that exists between the complex determination of processes and the complex construction of motivation/incitement to agency by multifarious outside sources of agency, a separation exists in the conception of necessity/determination and motivation/incitement.  Specifically, ANT insists on the role of actors as mediators of outside influences/sources of motivation to action.  Thus, actors are made to act, but only to the extent that they freely internalize particular influences that cause them to act in a particular way – individual actors are not openly effected by outside determinants.  The understanding of overdetermination advanced by this project, by contrast, omits the possibility of mediation.  Any apparent mediation of effects on an overdetermined process manifests the incomplete resolution of structural causation (i.e. of the determinate constitution of a particular locus by an infinity of processual effects across space-time).  This project rejects the ontological possibility of mediation for the same reason that it rejects the notion of the relative autonomy of particular structural levels, initially advanced by Althusser (Althusser and Balibar, 2009: 115), on the field of structured totality – this conception amounts to an attempt at theoretic resolution for the irreducible complexity of an overdetermined totality. 

Proceeding further, by insisting on the existence of an ontological background (i.e. structured totality) whose theorization shapes the processes through which theoretic knowledge is produced, determining that all theoretic knowledge must be partial and partisan, this project proceeds from a very different epistemological basis than ANT.  ANT theorists like Latour, Law, and Callon proceed from an epistemological approach that can, at best, be described as a kind of infinitely skeptical empiricism, forever holding out the hope for ever more objective articulations of empirical facts that can unravel the complexity of material existence but arguing that the theoretic production of objective truths, in either the natural or social sciences, is a continuously open and incomplete process.  Conceptually, Callon (2007) describes this kind of epistemological approach as performative, in the sense that theoretic statements, striving toward objectivity but never reaching it, participate fully as actors with other human and/or non-human actors in the acts of agency that they motivate.  Theory, thus, performs a particular reality in participation with other actors in which the influence of theory, as a component within the larger hybrid, is mediated through the mobilization of the other actors (e.g. Callon, 1986).  The outcomes of this reality and, thus, the objectivity of theory remain continuously uncertain because theory must mobilize all of the other actors to perform in a predicted manner.  Critically, ANT accounts may not conceive of the possibility of reaching closure in the production of objective “matters of fact” (Latour, 2007: 115)[10], but they nonetheless acknowledge as their goal to strive toward such illusive ends.  

My borrowing of Callon’s concept of performation, by contrast, seeks to acknowledge, with Callon, that theoretic processes shape the realities that they describe while substituting, for faith in the quest for objectivity through empirically-founded theoretic methods, hope that theoretic accounts can persuasively effect the actions of individuals exposed to theory in ways that achieve the partisan ends of the theory.  The difference between these two perspectives invariably leads back to the basic difference between mediation and overdetermination.  The effects of theoretic processes in an overdeterminist account are uncertain, not because they fail to mobilize other actors, but because the overdetermination of the reality that theory seeks to address is too irreducibly complex to permit the articulation of an objective account or an objective understanding of how theory does in fact effect it.  Thus, where ANT articulates networks as empirically traced connections between actors for the purpose of participating in a search for empirically-founded objective truths, this project seeks to articulate networks as theoretically conceived connections between processes in an effort to contribute to uncertain partisan projects of social transformation in an irreducibly complex material reality in which objectivity can never even be approached. 

It suffices to say in concluding this section that the theory of process networks that I have advanced does not belong to the body of Actor Network Theory.  On the other hand, particular themes advanced by ANT appear persuasive enough to merit integration into a larger understanding of spatio-temporality consistent with an overdeterminist, process-based ontology.  Specifically, the acknowledgement that integrated class processes, constituting specific class structures, involve dispersed, non-simultaneous spatio-temporal relations demands the development of an account to specify the particular spatio-temporal relations between processes of surplus production, appropriation/distribution, and receiving and to inquire into the formative significance of dispersion.  By insisting on the theoretic traceability of spatio-temporal connections and, above all, the necessity of transmission vehicles to connect all points, arguments emphasized strongly by ANT, Marxian class analytic accounts can develop new understandings on the construction of globally-extensive capitalist firms, interregional non-capitalist class structures, and non-class structures existing as conditions of existence to myriad, diverse class processes in multiple geographic locations.  For these reasons, it appeared imperative to situate my own theoretic approach and acknowledge its indebtedness to this other theoretic body, notwithstanding the clear differences that exist between ANT and overdeterminist ontology. 

---

[7] Following Law’s (2007) warnings on defining ANT, this body of theory should neither be characterized as a theory, per se, nor as uniform body of theoretically informed explanations on society.  Rather, it constitutes a loose body of approaches to analysis insistent on the empirical traceability of connections between actors through their acts of agency. 

[8] The best example of such hybrid agency is apparent in Callon’s (1986) account on the domestication of scallops for commercial harvesting.  He explains, for example, the possibility for contestation and betrayal of collective agency by individual actors (scientific/oceanographic researchers, institutional colleagues/research financiers, fisherman, scallops), through which the inability to mobilize one actor causes the network to fail. On Callon’s explanation, see also Murdoch (1997: 738-740).

[9] Law’s (2003: 5) characterization of translation as process and his argument that “actor-network theory assumes that social structure is a not a noun but a verb” muddies any clear delineation between Althusserian overdetermination and ANT in relation to process.  In this regard, my differentiation between the two ontologies cannot rest on the argument that ANT constitutes an ontology of materially static agents.    

[10] Latour (2007: 114-115) differentiates, in particular, between matters of fact and matters of concern, where the former describe non-human objects that lack the capacity for mediation and the latter describe non-human mediators.  In this respect, he introduces a differentiation between a first and second empiricism, the former concerned solely with matters of fact and the latter, associated with ANT, investigating matters of concern.  For Latour, there is no question that ANT accounts seek to be empirical and objective, but such accounts can simply never realize a premature unification (Ibid: 115) of matters of concern into matters of fact. 

Cities, Networks, and Globalization I

Chapter 2: Cities, Networks, and Globalization

The material to follow in this chapter fulfills key tasks of the larger project presented by this document.  First, it defines the city as a spatial form.  Second, it defines globalization in relation to the definition of the city.  Finally, it relates the city and globalization to class, preparatory to a consideration of class structural heterogeneity in chapter 3. 

In order to achieve the first of these tasks, I will address the basic problem of space in reference to the Althusserian overdeterminist ontology asserted in chapter 1.  The intention of this explanation is not to rewrite overdeterminist ontology in reference to space but to develop separate imageries of space on both the holistic, closed, determinate field of the structured totality of material processes and the partial, open, and contingent field of theory.  The imagery advanced in regard to the ontological field of theory will be that of the network, constituted as a piece-wise, partial assemblage of space.  It will be the purpose of the first section of this chapter to outline this dual ontological imagery of space and, especially, to define how networks will be understood in this project, comparing and contrasting this definition of networks to that of Actor Network Theory (ANT).  Proceeding beyond the basic effort to define the network, the second section will define the city as a spatial form constituted by networks and the third section will argue that globalization, as a constellation of processes defining distinct networks across space, is articulated through the space of cities. 

The remaining sections of this chapter will seek to elaborate further the distinctiveness of the city, as the urban spatial form, relative to non-urban forms.  The point, in this respect, is that the city is unique as a space for articulation of networked linkages between processes with globally extensive effects.  Moving beyond this particular uniqueness, however, the interrelations of processes in the city, within its hinterlands, and across globally extensive networks produce a particular developmental dynamic that I intend to label incubation.  In development and utilization of this concept, I mean to argue that the profusion of effects from processes internal and external to the city generate a dynamic of continuous economic redevelopment, in the sense described by Jacobs (1970: 125-129) in her conception of a supply-export reciprocating system and the creation of “new work” by cities more generally.  Asserting incubation in relation to the networked-defined spatial form of cities as the basis for development of new production processes in cities, the next two sections will articulate the consequences of incubation to class processes.  The intention, in this regard, will be to argue, on the one hand, that class processes articulate networks over diverse geographic ranges and, on the other hand, that the existence of geographically localized class processes in a given city incubates ranges of other production processes, extending class networks to other nodes internal to the city.  This consideration of the effects of incubation on class processes will constitute the starting point for elaboration of the principal hypothesis of this project that cities incubate heterogeneous class structures, in chapter 3. 

Concluding this brief introduction, the material contained in this chapter attempts to move from a high level of theoretic abstraction, from the field of structured totality, to successively higher levels of theoretically reconstructed concreteness evident in a definition of the city, as an assemblage of networked processes, including class processes.  I make no apologies for the level of abstraction advanced at the outset.  Rather, such a process follows, in part, Marx’s “method,” advanced in the introduction to A Contribution to the Critique of Political Economy (1993: 100-108).  That is to say, the methodology followed in this chapter, first, asserts the irreducible complexity of material existence, patterned theoretically by the concept of structured totality, with its own irreducibly complex spatio-temporality (universal space-time).  Beyond this irreducibility of structured totality, I seek to define a basic theoretic concept in the disconnected process with a disconnected spatio-temporality.  Finally, the approach reconstructs the connectedness of processes and the implications of connectedness in space-time (networked space-time).  In my view, the absence of such a theoretic process, linking Althusserian ontological concepts to spatio-temporality, would compromise the ability of this project to conceptualize either the network or the city by depriving these concepts of a solid connection to the Marxist/Althusserian ontology appropriated by this project as its basic theoretic framework.  For this reason, I regard the material contained here as constituting a critical stage in the larger conception of this project. 

Theorizing Space: Sites, Networks, and Nodes

The primary objectives of this section are to integrate space, or, more properly, spatio-temporality, into an overdeterminist Marxist/Althusserian ontology, and, with regard to the field of theory, to develop a conception of networks emanating from overdeterminist ontology.  In securing these objectives, I will, first, describe space-time on the field of structured totality, approaching spatio-temporality in reference to the concept of the site.  I will, then, describe spatio-temporality on the field of theory by pursuing the theoretic dislocation of the site from its context in structured totality.  The intention of this dislocation is to achieve the transformation of the site into a node, constituting the basic unit of a network.  The closing arguments of this section will advance a comparison of this approach to that of Actor Network Theory in order to more thoroughly illuminate the specific theoretic content of my approach to the network in relation to this more developed body of theory. 

Spatio-Temporality of Structured Totality: Sites

As outlined in chapter 1, structured totality constitutes the fundamental ontological concept in Althusserian ontology, describing the collection of all material processes, connected across multiple, hierarchical structures in which every process exists as a loci for determination by every other process, both internally and externally in relation to a given process. 

To proceed further in describing this overdetermination of processes, at each locus processual determinations from past and contemporaneous effects combine to produce a specific, determinate transformation of material existence.  I mean to construe the concept of “effect” utilized, in this regard, in the broadest possible terms, to include the particular transformations of material objects and those of contexts in which individual agents act or other processes otherwise occur.  Thus, the manifestations of use values emerging from a particular production processes in which raw materials, equipment, and labor power have been combined to produce something new may be construed as effects of the production process.  Additionally, emissions of chemical substances, incidental to the same production process, may also be viewed as effects of the production process, which, in turn, shape particular ecological processes through their transmission.  Lastly, the production process may impact myriad other political, economic, cultural, ecological, physical, chemical, biological, etc. processes in particular ways by means of other unspecified effects on the contexts in which these processes occur.  

Every combination of effects is specific, because each combines a unique set of determinations from different, unique sets of overdeterminants.  Some of these overdeterminants may arise from the explosive diffusion of effects of processes in the immediate past.  Others may arise from the accumulated effects of processes in a more distant past over an extended length of time.  Still others may be the effects of a process in the very distant past whose positive effects on the process awaited the appropriate combination of other, more recent effects to achieve the conjunctural realization of some particular outcome. 

These reflections begin to establish an understanding of the temporal dimension of processes in an overdetermined totality.  Time, in these terms, is relational.  As suggested briefly in chapter 1, my use of this terminology follows, in some degree, its use by theorists like Harvey, Massey, Amin, and other who identify the dimensionality of processes and/or things within material existence as a relationship to other such processes and/or things.  Harvey’s specification of the relational concept in regard to spatiality provides the clearest example of what I have in mind here.  Harvey characterizes relational space as a Leibnizian concept defining the dimensionality “contained in objects in the sense that an object can be said to exist only insofar as it contains and represents within itself relationships to other objects” (Harvey, 1973: 13).  More succinctly:

(T)he relational view of space holds there is no such thing as space or time outside of the processes that define them…Processes do not occur in space but define their own spatial frame.  The concept of space is embedded in or internal to process (Harvey, 2006B: 123).

 

The same sort of understanding of the relational concept can be applied with respect to the temporal dimension of processes within overdeterminist ontology.  Time exists as a set of relationships connecting a process to its internal and external overdeterminants, expressing a difference between processes occurring simultaneously and others occurring at successively farther distances into the past.  In this manner, time can only be gauged as a temporal distance between the manifestation of a process and some other process whose effects have overdetermined the first process.  The temporality of a process expresses its simultaneity or non-simultaneity with other processes that effect it as an outcome of overdetermination.  Thus, time is a dimension specific to each process, expressing the trajectory of a process in relation to all other processes.[1] 

Spatiality constitutes the second set of dimensions unique to each process.  Defined again in relational terms at least partly consistent with those of Harvey, spatiality describes the collocated or non-collocated existence of a process with its internal and external overdeterminants, in the limit, in simultaneity.  By extension of this relational logic, the notion of spatial dispersion or distance between processes can be defined in reference to the relative proximity of multiple processes, through which a process may be relatively closer to certain overdeterminants at a temporal instance than it is to others. 

By this understanding, temporal/historical ranges and spatial/geographic scales both exist as relational articulations, internalized to individual processes at specific loci.  Some overdeterminants of a process are collocated with and simultaneous to the process, while others are non-collocated (either internally contained or external, overlapping or entirely distinct and separated) but simultaneous to the process, still others are non-simultaneous but collocated, and still others are neither coextensive nor simultaneous.  The possibilities advanced here map the objective geography and history of a process in relation to its overdeterminants, objective in the sense that they conceptually situate the position of a particular process, at a point/moment in its trajectory, relative to every other process in structured totality.  The notion of such an objective geographic and historical position describes the concept of a site.[2] 

Sites have a geometry and a temporal instance/moment, encompassing all that is internal or organic to the process, including processes constituting the relative boundedness of the site itself.  For example, the class process of surplus labor production has a site, defined by the productive consumption of labor power, the spatial dimensions containing this transformation and the temporal moment in which the transformation occurs or does not occur.  This site is consistent with the notion of the “industrial site,” containing the simultaneous and collocated processes of surplus production and appropriation, as described by Ruccio et al (1991).  The site has boundaries, however tangible or intangible, that delimit the spatial dimensions of the process relative to what is external to it at each temporal instance/moment.  Internal to those boundaries, the class process of production contains certain overdeterminate processes, involving biological, physical, and chemical transformations (e.g. composites of chemically transformative processes changing raw materials into new use values) and mental, emotional, political, cultural, and economic processes whose outcomes effect the productive consumption of labor power. 

The site of the surplus production process contains the sites of many of its overdeterminants, but containment of these sites may be transitory.  Other processes may invade or escape the site of production at different moments, and the effects of these processes on the production process may change in relation to their containment or externalization.  For example, the sites of particular non-class social processes, involving workers engaged in a surplus production process and overdetermining the process, may be contained by the space of the production process at certain times and spill over or transport themselves to other sites at other times, with changes in location having effects on the production process.           

Extending this understanding of the site to encompass the aggregate, the spatio-temporality of structured totality, as the assemblage of all mutually constituted/mutually constitutive material processes, constitutes a seamless, universal, objective space-time, as the assemblage of all processual sites.  Such a universal space-time is decentered and devoid of any temporal baseline, because it constitutes an irreducibly complex aggregation of processes, each with its own relational centeredness (as individual loci), temporal baseline, and trajectory, defined by its relationship to all other processes.  In this sense, universal space-time can only be mapped in reference to individual processual sites – the spatio-temporality of structured totality, per se, cannot be mapped. 

More practically, this project asserts that the irreducible complexity implied by overdetermination precludes any objective mapping on the ontological field of structured totality.  To conceptualize such a mapping of individual processes would amount to the tracing of connections, in the transmission of effects, between processual sites across both infinitely expansive spatial and temporal dimensions.  Again, by asserting the unapproachable conditions for objective spatio-temporality, this project seeks to reinforce the performative character of geographic and historical theorization and analysis.  Mappings of space-time relationships between processes can only be partial in their articulation, and the purpose for engaging in such theorizations and analyses can only be to shape the way the reality of space-time is understood and, hence, lived.  Performative mappings do not belong to the objective and unapproachable background of totality but to the field of theory. 

Spatio-Temporality of Theory: Networks and Nodes

Abstracting from the irreducible complexity of structured totality and, thus, from the universal space-time it constitutes, the Marxian method implies that a theoretic ontology must isolate a simplified processual unit – a single locus for overdetermination stripped of its connections to processual overdeterminants – and reconstruct its constitutive linkages to other processes.[3]  In regard to space-time, the site, as a universally connected position of a given process in mutually constitutive relation to its overdeterminants, must be emptied of its spatio-temporality in order to reconstruct, in a piece-wise manner, the spatio-temporality of its connections to other processual sites. 

            In a spaceless-timeless ontology, the theoretic reconstruction of processual connections constitutes the articulation of structures, as series of processes defined by a particular, theoretically identified relationship between processes (e.g. the class structure).  In reference to space-time, reconstruction of the spatio-temporality of structures defines a spatio-temporal relationship between processes that I will label the network.  Networks, as they will be understood in this project, are fully theoretic constructs.  They arise from the theoretic abstraction of sites from structured totality and the subsequent (partial) reassembly of sites and their processes to articulate a particular geography and a particular temporal sequence or simultaneity, reflecting a structural relationship between the processes and processual sites contained by the network.  In this sense, the conception of networks used here remains at a relatively abstract, ontological level, encompassing any articulated spatio-temporal connection between material processes, not simply relationships defined by the presence of intervening physical infrastructures (e.g. an electrical grid or a mass transit system).[4]  The spatio-temporal connections between separate but functionally related chemically transformative processes in a photosynthesizing plant, thus, constitutes a network in the same way that the spatio-temporal connections between separate but interrelated financial processes involving multiple firms constitutes a different kind of network, constructed by a different theoretic process.

In order to emphasize the theoretic transformation of relations between sites in the articulation of the network relative to relations in structured totality, I will apply the concept of the node to describe particular sites contained by the network.  Nodes contain within their dimensions processes that are organic to the network as the spatio-temporal analog of a structure.  Nodes may also contain the sites of processes that are not connected theoretically to the processual structure defining the network.  For example, a production node in a network defined by a class structure (i.e. the site of production in a class structure articulated spatio-temporally as a network) contains a wide range of physical and social processual sites that do not define the class structure, even though these processes are overdeterminants of the class process of production.  The larger point, in this respect, is that the articulation of networks through the linking of nodes does not proceed from a general assumption of overdetermination between all processes (as on the field of structured totality), but necessitates the theoretic specification of linkages between processes and their spatio-temporal dimensions.  Networks may connect spatially dispersed nodes configured as archipelagos of processual sites, interspersed by processual sites wholly disconnected from the network as it is theoretically articulated in abstraction from the infinity of processual overdeterminants by which it is constituted.  

            With specific regard to the spatiality of networks and their respective nodes, some nodes will be relatively stationary over given periods of time while others are relatively mobile.[5]  Considered, again, strictly with respect to the spatial dimension, this must be the case if the processes defining the network are not spatially collocated.  The spatial dispersion of network processes implies that some processes, internal to the network, must connect relatively stationary and spatially separated processes.  A type of mobile node must exist to facilitate the transmission of effects between stationary nodes, a node defined by its processual role in the transmission of effects.  Extending this logic to the temporal dimensions of a network, if processes in a network are collocated but not simultaneous or sequential (i.e. separated by extended temporal ranges), then the network requires a node defined by its processual role in the transmission of effects between separated temporal instances.  The point is the same for networks defined strictly in reference to spatiality, strictly in reference to temporality, or in reference to some complex relationship of space and time: stationary nodes in a dispersed processual relationship require a vehicle for the transmission of specified effects across spaces or times not otherwise articulated within the network.

            A relatively simple relationship between a set of processes may illustrate the understanding that I mean to convey about relatively stationary and relatively mobile nodes.  Visual observation of objects by a human observer involves an integrated set of processes, characterized by spatial and temporal dispersion, constituting a network.  In spatial terms, the process involves two relatively stationary nodes: the object (constituted by a set of internal and external physical and social processes that overdetermine its existence as an object in continuous material transformation) and the human observer (constituted by a set of internal and external physical and social processes that overdetermine its existence as a human individual in continuous material transformation).  The spatial distance between the two relatively stationary nodes constitutes a pathway for the transmission of effects, but this pathway is constituted by processes (e.g. chemical transformation of molecules in air) incidental to the theoretically defined structure of processes in visual observation.  What is needed is a processual vehicle to connect the object to the observer.  This processual vehicle arises from the reflection of visible light waves off the surface of the object.  The process of visible light radiation, capturing the effects of processes constituting the object, thus, constitutes a transmission process between the relatively stationary nodes in the network.  The spatio-temporality of light particle transmission from the object, varying in accordance with the speed of light waves in continuous reaction with the physical and chemical processes of the pathway, exists as a mobile node in the network, carrying effects between stationary points.  A simple articulation of the network, abstracting from the complexities involved in optical processes or interpretation of visually accumulated information, is completed in the absorption of visible light waves by the observer’s eye. 

In order to further explain the meaning of the network in relation to the particular kinds of processes with which this project and Marxian theory, as a whole, are especially concerned, the class structure defined in chapter 1 provides an important example.  Again, the understanding of class structures followed within this project involves the integration of a particular set of class processes (the production, appropriation, distribution, and receiving of surplus labor) and the connections to non-class processes securing conditions of existence for the production and appropriation of surplus labor.  Class processes contained by a given class structure may take place at a variety of spatial locations and temporal instances.  The spatio-temporal relationships between these organic, relatively stationary processes, connected by mobile transmission processes, define a particular network, the class network. 

To present an example, a particular commodity-producing firm, defined by a class structure, produces commodities at a given spatial location over a particular period in time.  The production process involves the physical transformation of certain raw material use values into a different set of use values, but it additionally involves the production of surplus value through the exploitation of labor power to produce an additional realizable value embodied in finished commodities exceeding the exchange value of the tools, equipment, raw materials, and labor power utilized in production.  Again, following Ruccio et al (1991), this site is simultaneously the site at which the firm appropriates surplus values, contained as values congealed in a mass of finished use values.  The realization of these surplus values in a monetary form occurs through market exchange.  It is conceivable that this realization process occurs at the same site as the site of production/appropriation, but I will assume, for my purposes, that the commodities must be relocated to a different site for market exchange.  Upon realizing the surplus values contained by the commodities through exchange, the appropriator/distributors of the firm receive the realized surplus values, perhaps with a deduction reflecting a surplus distribution to secure the realization process (e.g. to a retailer).  This receipt of surplus value may, again, occur at any site, but I will assume that it occurs at a site that is distinct either from the site of production/appropriation or the site of market exchange/realization, necessitating some forms of financial transmission to the site of initial receipt and distribution.   

The combination of these processes conveys a particular temporal sequence of events (production/appropriation – realization – receiving) and a particular geography (site of production/appropriation - site of realization – site of initial receiving).  It also involves transmission processes with relatively mobile sites (e.g. site of physical commodity transportation from the production/appropriation site to the site of realization; site of electronic transmission of financial information from the site of realization to the site of receiving).  By articulating all of these processes as an integrated structure of processes (i.e. a class structure) and applying a set of spatio-temporal relations (simultaneity, sequence, collocation, proximity, dispersion) between the sites of these processes, I articulate the structure as a network and redefine the sites as nodes within the network. 

Elaborating, I can further extend the network by adding other conditions of existence for the production and appropriation of surplus value (i.e. other than market realization).[6]  Such processes as the provision of financing to undertake the production process may, for example, involve processes (e.g. lending by financial institutions) wholly separated from the sites of production/appropriation, realization, and initial receiving of surplus value by the appropriator/distributors.  The financial process is connected, structurally, to the class structure through the distribution and receiving of surplus value in the form of interest payments.  The distribution and receiving of these payments may have their own spatio-temporalities, separated spatially and/or temporally from the appropriation of surplus value, and the receiving of surplus values in the form of interest payments by financial institutions may be spatially and/or temporally separated from other processes involved in financial lending to secure the financial conditions of existence for the firm to undertake production.

Elaborating still further, the transmission processes within the network may connect to myriad other processes, otherwise disconnected from the network, if these transmission processes result in certain metaphorical leaks.  That is to say, if the transmission process between stationary nodes results in non-negligible effects to the pathways along which transmissions take place, then the class processes in the network will be affected in ways that can be explicitly theorized.  To provide a very tangible example, if commodity transportation processes between a production/appropriation node and an exchange node by means of a waterborne freight process results in ecological damage (e.g. contamination of water from an actual fuel leak), then the class network will be shaped by the interaction of economic, political, and cultural processes resulting from such an event.  Every process connected to the network as a result of the event possesses its own spatio-temporality, implying that for every such event, new stationary and mobile transmission nodes must be connected to the network.   

In specifying the spatio-temporal relationships contained by this class network, the spatio-temporality of all stationary nodes has to be specified and mobile nodes for the transmission of effects along pathways must also be specified in order to theoretically articulate a definitive relationship between these processes that is performative of the reality that Marxian theory means to convey.  For every new processual connection specified, a new, stationary node is added to the network, connected by some transmission node.  In this manner, the construction of the class network is piecewise.  It does not saturate geographies with explosively diffusive connections, but reconstructs spatio-temporal relationships between particular processes one connection at a time. 

            To summarize, the network concept, as I have presented it, describes a theoretic articulation of sites, extracted from structured totality, containing processes whose integration defines a structure.  The theoretic repositioning of sites in this articulation transforms each site into a node.  Spatial and/or temporal dispersion of processes within networks necessitates that some nodes will be relatively stationary while other will be relatively mobile, facilitating connections across spatial or temporal pathways between the stationary nodes of the network.  This understanding seeks to preserve the processual understanding of material existence advanced by Marxian theory and the definition of space-time as the dimensions of processes.  If all space-time exists as a dimension of processes and structurally integrated processes exist in a spatially dispersed organization, with intervening spaces constituted by processes unrelated in theory, then the connections between structurally related processes must require some means to facilitate the transmission of effects between sites/nodes. 

---

[1] This understanding of time, as a dimension of process in an overdeterminist ontology, corresponds to Althusser’s assertion of heterogeneous specificities of the historical temporality of different “levels” within a Marxist totality (Althusser and Balibar, 2009: 110-111).

[2]  This concept of the site corresponds, in part, to the site concept advanced by Resnick and Wolff (1987: 231-232), defined as “a location or grouping of specific relationships that each comprise a particular subset of social and natural processes.”  The definition advanced by this project, on the other hand, seeks to advance a more specific spatio-temporal definition of the site arising from a relational logic of space-time.  In this manner, the site concept introduced here is read as a geographically and historically specified and bounded location defined by the spatio-temporality of processes both internal and external to a set of boundaries defined by the process. 

[3] Such a locus might be characterized, as in Resnick and Wolff (1987: 25-30), as an entry point, in contrast to an ontologically prioritized causal essence.

[4] This level of theoretic abstraction, further, contrasts with the use of the network concept by Castells (2000A:163-215), who uses it describe information flows in globally extensive capitalist firms, and by Hardt and Negri, who variously employ the concept in the same manner as Castells (Hardt and Negri, 2001: 294-300), to describe the expansiveness of the U.S. imperial sovereignty (Ibid, 164-167), and to describe the organizational styles of al-Qaeda, the Zapatistas, and the “movement of movements” against capitalist globalization (2004: 54-57, 79-91).  

[5] The stationary property of certain network nodes is only “relative” because the geometry and geographic positioning of processual space is continuously relational (i.e. the space of every process is continuously defined by its changing position relative to other processes).  In this sense, the network presupposes the capacity of theory to freeze a temporal moment or a sequence of moments in order to articulate the spatiality of a set of relations between a specified set of processes in a structure.   

[6] Resnick and Wolff (1987: 174-180) elaborate a list of such subsumed class processes, involving the distribution of surplus values to non-productive agents (i.e. those not producing realizable surplus values but performing other necessary work) inside and outside of a firm.

Introducing Marxism and the City in Globalization VII (Conclusion)

Thesis: The Heterogeneity of Class Structures in Globalizing Cities

The first section of this chapter concluded with the claim that the central commonality evident in the various approaches to the city in Marxist and Marxian-inspired theories is that the contemporary city is conceived as a capitalist city.  Numerous accounts advance such a conclusion quite explicitly.  For example, Harvey argues, in reference to the variety of technological changes reducing the temporal requirements of movement across space (i.e. space-time compression), that:

These new technological and organizational possibilities have all been produced under the impulsions of a capitalist mode of production with its hegemonic military-industrial-financial interests.  For this reason, I believe it is not only useful to think of but also important to recognize that we are all embroiled in a global process of capitalist urbanization or uneven spatio-temporal development even in those countries that have nominally at least sought a noncapitalistic path of development and a noncapitalist urban form (1996: 414, italics in original).   

The implication of this reasoning is that capital accumulation in diverse production processes and the particular effects of technological development arising from such processes imprint the image of a globally expansive capitalist mode of production upon all social contexts with which they come into contact, notwithstanding the efforts of policymakers to displace axiomatic dependence on capitalism.  In reference to the city, as a spatial form, the impact of capital accumulation, both in the development of urban built form and the integration of technologies into the everyday lives of urban residents, ensures that every city in the age of a specifically capitalist globalization, marked by the insistent penetration of capital accumulation into new geographies, must be a capitalist city.  Harvey’s characterization of the relationship of capitalism to urbanization can be described as capitalocentric, using Gibson-Graham’s (2006A) terminology.     

            The capitalocentric characteristics evident in the larger body of Marxist and Marxian-inspired theories on the city manifest diverse forms, introducing divergent consequences for the relationship between theory and urban practice.  As suggested in the previous section, theories focusing on the relationship of class struggle to the development of urban social movements manifest a particular reduction of class to the binary struggle of unitary working and capitalist classes, a binary constituted by the image of a capitalist mode of production that is singular and coextensive with economic processes such that every form of workplace involved in commodity production can be labeled capitalist.  The commitment of agency-oriented theorists to ontologically privilege individual agency relative to structural determination implies that such discourses do not, however, typically go as far in expanding the sphere of effects from capitalism as more structurally oriented theorists like Harvey, the regional economy theorists, or the control and coordination (global and world city) theorists. 

To varying degrees, structurally oriented approaches to the city expand the social effects of capitalist processes to suggest, in many cases, a kind of uni-directional determination of political, cultural, and ecological processes by the economic, at a minimum arguing that state policies in urban planning or collective consumption and the development of alternative non-class cultural self-identifications must be functional to capitalist development and reproduction.  Further, reflecting on the position of the city relative to other geographic scales, the reduction of the city to a limited set of political/institutional roles by the regional economy theorists, alluded to in the first section in reference to Scott et al (2008), introduces a kind of inter-scalar economic determination of the local (i.e. the city) by the regional and/or the global, in which the latter, higher scales articulate restructuring processes by industrial capitalism.  In this sense, capitalism, as the economic, is uniquely capable of transcending the limitations of local space and even regional and national space to become global.  Thus, the city becomes a captive of regional and/or global economic processes that are unambiguously capitalist. 

Structurally oriented manifestations of capitalocentrism and the reduction of class to the working/capitalist class binary ultimately reflect the utilization of definitions for capitalism that are sufficient broad to subsume not only a range of effects on non-economic processes, but to subsume the totality of economic processes and to reduce the possibility or actual presence of alternative organizations of surplus labor to relatively minor variations on a common capitalist logic of commodity production and/or market exchange under the motive of profit.  Capitalism, thus, completely encompasses the horizon of feasible economic structures and this coextensive identity of the economic with capitalism constitutes the hegemonic influence of capitalism on political, cultural, and physical/ecological processes.  Consequently, socially transformative projects, contesting the effects of capitalist economic processes, confront an expressive totality of social processes, all arrayed to reproduce the hegemony of capitalism.  For this reason, the sorts of capitalocentric discourse represented in most of the aforementioned Marxian theories of the city impose on socially transformative urban movements, especially those anchored on the historically assigned agency of the working class, the impossible task of overthrowing, in its entirety, a globally integrated system from the standpoint of the local/the city.         

Consistent with its theoretic bases in the Marxian theory of class, Althusserian overdeterminist ontology, and performative epistemology, this project rejects this capitalocentric imagery.  Elaborating, overdeterminist ontology holds that capitalist class processes are mutually constituted by and mutually constitutive of all other social and physical processes, implying that no uni-directional determinism is conceivable.  As such, the structured totality of processes containing capitalist class processes manifests complexity, irreducible to the simple expression of capitalist dominance in all its processual parts. 

Addressing the geographic/scalar imagery of global capitalist dominance and ontological captivity of cities relative to the regional or the global scale, this project will advance a theory of networks in chapter 2 with the intention of reframing the geography of the city in relation to material processes.  This network imagery will insist that spatial scales are constituted as piecewise constructs of relatively local processes, local in the sense that every process has a definite positionality and boundedness relative to other processes for any given moment in time.  The global scale of social processes exists as a piecewise combination of local processes whose effects diffuse explosively but unevenly along complex but fragile networks.  Consistent with overdeterminist ontology, such networks must, likewise, facilitate multi-directional effects.  Thus, the effects of globally dispersed capitalist class processes on a particular city cannot be understood, on the one hand, as uni-directional (from transnational capitalist firms to individual cities) or, on the other hand, as emanating from a unitary global agency (from the global to the local) rather than as an assemblage of networked but dispersed local processes.  Such an understanding of connections, not between the local and the global but between diverse locals across networks that facilitate multidirectional linkages, imposes the sort of relational logic discussed earlier in which every process is understood to define its own scale and agents interact simultaneously through multiple local, regional, and globally-extensive networks defined by processes.

Finally, and most importantly for the hypothesis that I mean to initially advance here and subsequently to elaborate in chapter 3, this project contests the ambiguity with which capitalism has been defined and ascribed a substantive influence on a wide range of social processes in the theoretic approaches heretofore described.  Capitalism, as it is variously defined implicitly or explicitly, describes market exchange processes, the integration, expansion, and transformation of social divisions of labor with respect to production processes and geographic regions, the supremacy of the profit motive in driving entrepreneurship, or the evolving technical, bureaucratic, and geographic organization of enterprises.  If, in some sense, all of these potential definitions of capitalism describe economic processes, they are all too broad and diffuse to convey the unique insights that Marxian theory has to offer on class in general or on capitalism in particular.  They truncate the theoretic spaces for identification of economic non-capitalisms.

The first step in moving beyond this capitalocentric logic involves providing a definition of capitalism consistent with Althusserian overdeterminism.  Applying, again, the definitions of class structure types conveyed by Resnick and Wolff (1987), capitalism constitutes a particular exploitative class structure in which the producers of surplus labor supply their capacity to do work to a capitalist appropriator/distributor as commodity labor power in market exchange.  In exchange for wage-based compensation, producers produce a value in excess of the value of the labor compensation/wages (a surplus value), contained in a mass of commodities and realizable through exchange.  Upon realization of the value contained in the commodities, the capitalist appropriator/distributor determines the necessary distributions that must be undertaken in order to secure conditions of existence of the surplus production process so that it can be successfully reproduced over time. 

This modest definition describes and delimits capitalism as a class structure in such a way that certain class structures, existing simultaneously to capitalism cannot be described as capitalist.  Specifically, surplus production processes in which the producer and appropriator/distributor are identical and singular describe non-capitalist ancient class structures.  Surplus production processes in which the producers and appropriator/distributors are identical and plural describe non-capitalist communist class structures.  In both of these cases, means for the compensation of labor power, the realization of values, and the technical organization of production processes are incidental to the organization of surplus labor that defines the class structure.  Thus, ancient and communist class structures may produce commodities for market exchange and operate as firms, in a legal and proprietary sense, but market and legal/proprietary processes do not render them capitalist. 

By utilizing a definition of capitalism that relates it, as a class structure, to a very specific form in the organization of surplus labor, I am able to advance the outlines of a hypothesis.  Cities, as spatial forms, contain diverse ranges of social processes, organized ontologically into structures of related processes and spatially and temporally linked as networks.  Some of these networks are contained within the city and others extend longer distances beyond the variously defined boundaries of the city.  Some of these social processes are class processes.  Among the class processes, some are capitalist class processes.  On the other hand, the class processes and the structures and networks that they define are not exclusively capitalist.  There are slave, feudal, ancient, and communist processes as well.  All of these processes operate simultaneously.  Their structures are also typically interlinked and their spatial networks overlap.  In addition, capitalist class structures are not exclusive in producing networks that extend beyond the boundaries of a city.  Other types of class structures produce their own interregional class networks and non-class processes produce interregional non-class process networks (e.g. ethnic-cultural processes, political processes, etc.).  These preliminary assertions mean to posit the city as a spatial form existing as a piece-wise ensemble of processes, structures, and, spatio-temporally, networks and to argue suggestively that the existence of heterogeneous class structures in cities is important in shaping the city as something other than a purely capitalist city.  That is to say, the existence of non-capitalist class structures, especially non-exploitative communist class structures, shapes the economic, political, cultural, and ecological processes of the city. 

Further, in conformity with a performative epistemological framework, by positing the existence of non-capitalist class structures in cities, I mean to reinforce the potentiality for cities to nurture such structures.  In this sense, the theorization of non-capitalist structures as actually existing entities in cities and theorization of conditions of existence for such structures does not simply generate an empirically testable hypothesis for examination of cities, in the sense that Castells’ theorization of the relationship between collective consumption and urban social movements had motivated his empirical scholarship of the late 1970s.  Rather, this project seeks to contribute to a continuously evolving Marxian strategy to develop communist class structures in cities as practical alternatives to capitalism. 

            In, thus, departing from the capitalocentric logic of other Marxian approaches to the city, this account means to take Katznelson’s advice that Marxism indeed needs to exhibit a greater degree of modesty in approaching its revolutionary aims.  In contrast to Katznelson’s approach, however, the sort of modesty that this project seeks to cultivate concerns the scale of communist society envisioned by theory.  The sort of communisms envisioned by this project possess no pretension of overthrowing en masse capitalist class structures and all of the non-class structures supporting the reproduction of such structures.  On the contrary, this project understands communist revolution as a continuous and tenuous piece-wise endeavor of collective entrepreneurship and class and non-class network building in which cities play an important role.  For that reason, this project seeks to situate the city, as an interregionally connective space, in relation to class and to advance a more modest appraisal of the possibilities for a more modest communism than that advanced, implicitly or explicitly, in alternative Marxist approaches.

Introducing Marxism and the City In Globalization VI

Class and Class Structure (Continued)

Structured Totalities in Althusserian Overdeterminist Ontology

The conception of class and class structure introduced above alludes to a broader understanding, embodied in Althusser’s conception of ontological overdetermination, that class exists as one process among an infinity of mutually constitutive processes, which, together, define the totality of material existence (i.e. the universe of processes and substances/relationships/practices).  By concentrating on processes, this ontological framework emphasizes the centrality of transformative change to Marxian theory.  It asserts, moreover, that a seamless connection exists between all processes defining material existence.[9]  In this sense, the transformative effects of any process must be manifest as an effect on all other processes, rendering all processes, including class, loci for determination by all other processes.

            Understood in these terms, material existence, as a single structured totality in continuous transformation, is constituted by the processual transformation of all its parts, and such transformations are structured by the positionality of every process relative to every other process as loci for overdetermination.  Such an ontological imagery is holistic, in the sense that it encompasses material existence as a seamless whole, and structuralist, in the sense that the transformation of each process is completely determined by its positionality relative to other processes.[10]  On the other hand, as Althusser’s conception of ontological overdetermination will be understood in this project, the structured totality exists as a piecewise construct of processes, each with its own overdetermined trajectory in which no single process can explain the continuous transformation of the totality.  No process manifests autonomy, relative or otherwise, in relation to other processes contained by the totality.  Such a totality cannot, therefore, approximate an expressive totality of the Hegelian type (Althusser, 1970: 203-204) in which the evolution of the totality manifests the effects of a single, overpowering dynamic, amenable to relatively simple analyses through which the essence of the totality, expressed in all of its parts, can be revealed either by theory or by empirical analysis.  It should be clear, in this manner, that there can be no question of rigid determination of social processes by the economic in the last instance (Althusser, 1970: 111-113).[11]   

More importantly, the irresolvable complexity embodied in the overdetermination of each process ensures that no theory or analysis can produce objectively true accounts on any process.  This assertion of ontological overdetermination necessitates a reconsideration of criteria on the efficacy of theory beyond the criteria of objective validity.  In place of validity criteria, this conception of Marxian theory accepts, as principle, that theory and analysis can only present partial and partisan accounts on material existence.[12]  This understanding of the relationship between structured totality and theoretic processes implies that Althusserian Marxist ontology overlays two fields: the holistic, closed, determinate field of the structured totality, and the partial, open, and contingent field constructed through the theoretic articulation of processes.[13]    

Approaching theory with the greater degree of modesty compelled by this dualistic ontology, Marxian theory advances a standard of efficacy properly labeled performative, in a sense similar to that applied by Callon (2007).  That is to say, as material processes contained by the structured totality, theoretic processes actively shape their objects in reference to the perceived goals embodied in the theorist’s perspective.  Such a position concedes the role of the theorist as agent, participating in the processual transformation of material existence, and acknowledges the role of theory as a constituent process in material existence and, hence, a processual over-determinant of agency. 

            Recognizing this understanding of the performative nature of the Marxian theory of class advanced in this section in relation to the complexity of Althusser’s conception of structured totality, certain implications need to be highlighted.  First, Marxian class theory denies the centrality of the class processes as an essential cause of social reality.  In this sense, notwithstanding its grounding within a holistic and structuralist ontology, the Marxian theory of class is anti-essentialist.  Rather than asserting the ultimate centrality of class processes as an explanation for the importance of this project, the purpose for prioritizing class in a Marxist theory of cities in globalization pertains to the perceived impact such a theorization might have in performing a different reality, closer to the partisan goals of Marxian theory.  Further, in proceeding, my approach to cities in relation to class processes must also recognize the theoretic character of spatiality and spatial forms, like the city, implying that all aspects of this theorization seek to perform a reality, articulating connections between class and non-class processes to produce a complex ensemble of spatial forms defined as the city or, more concretely, the city in the present period of globalization.  As such, neither the city nor globalization can be taken for granted as realities existing prior to theoretic processes, like this one, that seek to define them in order to transform them.  Among the tasks of chapter 2 will, therefore, be defining the city and globalization in relation to this Marxian theory of class.

Comparing the Effects of Alternative Class Concepts in Theorizing Cities[14]

Reflecting a differentiation within Marxian theory, writ large, the approaches to theorization of the city introduced in the previous section generally acknowledge the importance of class but advance divergent conceptions of class, reflecting, in turn, ontological differences in the prioritization of structure and agency.  These differences draw distinctions between the various aforementioned approaches to the city in Marxian theory, but they are most apparent in the approach identifying cities as milieux for the development of urban social movements.

For example, Castells, in his writings of the late 1970s and early 1980s (1977; 1978; 1983), and Katznelson (1981; 1992) prioritize the role of agency in class conflict, examining, in this regard, the importance of agency by class-defined organizations in the development and outcomes of urban social movements.  For both of these theorists, class describes a social grouping, defined in reference to property ownership or income or skill differentiation and at least potentially capable of organizing in defense of common interests.  Class analysis, therefore, concerns examination of the processes through which individuals, unified by common working class characteristics, construct organizations and act in support of class-defined goals, within or outside of commodity production processes in capitalist workplaces.  In this respect, Castells’ work published after The Urban Question seeks to identify working class-based movements contesting the conditions for collective consumption as a component in the reproduction of labor power.  Katznelson’s approach goes farther in attempting to link structural determinants of class identity through four separate levels to class-based agency in urban social movements (1986; 1992).  In both cases, at stake is the linking of structural characteristics and structurally defined “class interests” to personal, individual identification with class and activity on the behalf of class interests.  For his part, Katznelson asserts most strongly, in concurrence with other theorists writing in the Marxian tradition (e.g. Aronowitz, 1992), that the perennial absence of a revolutionary collective agent in the history of Western capitalism suggests that Marxism should assume a more modest position in assessing the revolutionary potential of the industrial working class and, further, its capacity to support the development of class conscious urban social movements.   

The characteristic determinants of individual embodiment within the working class and of working class interests in these approaches prioritizing agency appear to be objective (i.e. class in itself) while identification by individuals with these characteristics, or those of gender, ethnicity, race, nation, or any other categorization, appears to be subjective (i.e. class for itself) and individual self-identification represents an open space to be filled by class or by any other discursive source of self-identification.  The contingent nature of working class self-identification and collective organization on the behalf of working class interests, for these theorists, implies that the capacity of Marxian theory to make useful statements about agency by urban residents is strictly limited.  Castells (1983: 296-300), for his part, goes as far as to conclude that Marxism, interpreted as a body of structuralist theory on the development of the forces of production through which working class movements enjoy no liberty to demonstrate actual agency, has nothing meaningful to say about the development of urban social movements.

            For Harvey, by contrast, urban analyses remain continuously at a more abstract level in which classes, understood as representative of structurally defined roles, follow structural imperatives.  Thus, fractions of the capitalist class pursue the de-valorization of urban infrastructures as a means of reproducing the possibilities for profit through investment in the secondary circuit.  Other fractions of the capitalist class (e.g. landlords) may possess an interest in maintaining the values of older urban infrastructure, but such contradictory individual interests are resolved on the behalf of the larger class interest through the exercise of power by dominant class fractions (e.g. finance capitalists), possibly with the assistance of the state.  In this respect, Harvey does not espouse a need to link structural imperatives to individual identifications because, for purposes of his analyses, structurally defined capitalist class interests drive urban economic processes, irrespective of any conscious identification by individual capitalists of a collective capitalist class interest.  A class analysis of cities, thus, implies the articulation of an account on how capital accumulation, driven by the efforts of the capitalist class, as a whole, to ameliorate falling rates of profit in commodity production, shapes the built environments of cities.  Again, for Harvey, the characteristics of the working and capitalist classes and their respective class interests are objective, but the theorization of an open space for self-identification in reference to class interests is, in itself, irrelevant because structurally defined imperatives, particularly those of capital (i.e. the anthropomorphized embodiment of the capitalist class as a whole), prevail.  Further, in contexts where self-identification of workers with the working class and its collective interests is impeded by alternative self-identifications (i.e. identity politics), Harvey (1996) interprets such alternative self-identifications as ideological, confused, and, ultimately, irrational products of postmodern culture.

            A comparison of these agency oriented and structurally oriented approaches to class and that of the overdeterminist definition of class introduced in this section, again, returns, in part, to the problem of the structure-agency binary in Marxian theory.  In relation to the agency-oriented approaches to urban social movements identified above, my overdeterminist approach shares a rejection of any axiomatic linkage between objective class characteristics (class-in-itself) and subjective recognition/self-identification and organization around class interests (class-for-itself).  My approach, further, recognizes the multiplicity of cross cutting, contradictory sources of self-identification, like class, gender, race, ethnicity, adding to these sources those emanating from contradictory class-positions held simultaneously by particular individuals.  Finally, this project rejects, with the agency theorists, simple determination of human agency by essential macro-level social causes (e.g. simple determination of human agency by technological change).  Among the more structurally oriented accounts to the city, this project accepts the larger implication of Harvey’s analysis that the multiplicity of effects from social processes, including but not excluded to the effects of capital accumulation, constitute structural determinants of the agency of individuals. 

            On the other hand, this project most emphatically rejects conceptions of class as a social grouping, central to the aforementioned accounts on both sides of the agency-structure divide.  It affirms a conception of class positionality relative to the organization of surplus labor.  The reasons for this position are multifold.  First, by identifying class in reference to surplus labor, the range of conflictual relations between holders of class positions expands to encompass conflicts over both labor power compensation/rates of exploitation and surplus labor distribution to secure non-class conditions of existence.  The range of potential groupings defined in relation to shared class positionalities and the potentiality for individuals to occupy multiple, contradictory class positionalities produces an imagery irreducible theoretically to a binary conflict between unitary, aggregate working and capitalist classes. 

By contrast, in defining class as a social grouping, the agency and structurally oriented approaches engage in a constricted, yet polarizing, debate over whether the working class manifests the freedom to act as an intentional collective agent or whether its agency is rigorously determined by economic processes against which any deviations constitute irrational behavior.  The presumptive existence of classes as groups, in general, and of a working class, in particular, with a particular, defined revolutionary role against a capitalist ruling class, thus, completely determines the terms of the debate and its contradictory tensions for theorists, like Katznelson, who seek to maintain the relevance of structural determinants while likewise prioritizing the autonomy of working class agents in urban contexts to act on them.  By denying class this meaning and, consequently, denying an axiomatic revolutionary role to a presumptively existent entity called the working class, my approach means to avoid such an ontological debate by reframing the larger transformative purposes of a Marxian theory of the city around the organization of surplus labor. 

Moreover, by insisting on the epistemological performativity of Marxian class analytic interventions into the city, my approach seeks to reframe the role of theory, per se, relative to what appears to both the agency and structurally oriented theorists.  Both sets of theorists accept a realist epistemological position (Gibson-Graham, 2006A), implying that theories and empirical analyses exist in neutral separation from their objects from which they can advance objective accounts of the reality of their objects.  This identification of the role of theory and analysis, in turn, shapes the ontological approaches embodied by diverse theorists.  For the agency-oriented theorists, Marxian theory represents a tool to elucidate the capacity and/or will of the working class to engage in revolutionary struggle against capitalism in reference to objective historical/empirical evidence.  Beyond its capacity to explain the agency of collective urban working class agents as a class, Marxian theory has no other functional utility.  The centrality accorded to empirical evidence, in this regard, prompts Castells to reject the theoretic insights of Marxism with respect to urban social movements.  By contrast, structurally oriented theorists like Harvey and Jameson represent Marxian theory as an objective body of rationally constructed insights against which the actions of working class agents can be adjudged for irrational nonconformity with structurally determined working class interests.  Such a position shapes the rejection of alternative sources of gender, racial, or locality-based self-identifications by Harvey as instances of ideological postmodernism. 

In contrast to both these realist perspectives on the role of Marxian theory, my approach understands Marxian theory and class analysis, in particular, as active participants in the overdetermination of the realities it theorizes.  In this sense, class analysis neither consists of a set of hypothetical assumptions to be verified against empirical evidence for validation nor does it consist of an a priori body of universal, rationally conceived truths asserting primacy against other merely partial and subjective bodies of social knowledge.  Rather, I understand the position of Marxian theory, together with that of every body of theoretic/analytic knowledge, as intensely practical in shaping the way its object is lived and resolutely partisan in acknowledging its political character in contesting particular other theoretic understandings of its object.  Marxian theory, as a body of knowledge, thus, constitutes itself as a performative argument, oriented toward the elucidation of class and engaged in a political struggle to shape reality by shaping the way its audience understands reality.  By affirming an active and socially engaged role for theory as agency and as structural/processual over-determinant of agency (e.g. the agency of individuals occupying multiple contradictory class and non-class positions), this approach rejects the existence of the irresolvable structure-agency binary shaped by realist epistemological approaches.

Acknowledging, at a relatively high level of abstraction, that the understanding of Marxian theory and class analysis embodied in this project rejects the identification of class as a social grouping, the binary alternatives of structural and agency-oriented theories, and the varieties of epistemological realism accepted broadly by the theorists discussed in the previous section, with the possible exception of Lefebvre, a more concrete assertion of a unique agenda, differentiating this approach from its theoretic predecessors as a Marxist theorization of the city, remains to be advanced in the concluding section of this chapter.    

---

[9] This assertion of a seamless connectedness between processes conflicts, in part, with Massey’s (2005:40-42) reading of Althusser’s critique of Hegelian totality and, in particular, the capacity to isolate “essential sections” in the Hegelian conception of time (Althusser and Balibar, 2009: 105).  Massey’s purpose is to demonstrate how Althusser’s critique facilitates, to some degree, a politics of space, relative to historical time, by asserting the relative autonomy and historical development of levels within a Marxian totality, implying heterogeneous ontological development of processes across space.  Massey contrasts this with the closed systems of structuralist synchrony and the Hegelian essential cross-section.  In contrast to Massey’s reading, my interpretation of overdetermination precludes both the relative autonomy of structural levels and the ontological openness of the totality while emphasizing, on the other hand, the decentered, irreducibly complex nature of the structured totality.

[10] Following Althusser (1976: 126-131), the intention in labeling theorization of a Marxist structured totality as “structuralist” is not to confer on Marxian theory any claim to analyze the formal trajectory of processes. On the contrary, Marxist ontology presupposes the existence of a totality within which every process exists in contradictory relation to every other process and that Marxism “affirms the primacy of contradiction over the process” (Ibid, 1976: 130).  The complexity implied by overdetermined contradiction makes any pure theoretic structuralism, defined in reference to objective formal analyses of processual relations, impossible.  The critical issue that separates Althusserian Marxism from structuralisms is, thus, the relationship between a holistic ontological background and the necessarily anti-essentialist theoretic argumentation that approaches it.     

[11] In this respect, this project accepts the critique of Althusser’s conception of structure “articulated in dominance” offered by Resnick and Wolff (1987: 93-94) arguing that Althusser failed to definitively break with the economic determinism implied by determination “in the last instance.”  The conception of ontological overdetermination employed in this project conforms more closely to Resnick and Wolff’s revision of Althusserian Marxist ontology to incorporate a more thoroughly anti-essentialist meaning.    

[12] By implication, the epistemological standpoint of Marxist theory is relativist in the sense that theory embodies a particular, overdetermined perspective on its object and engages in struggle against alternative theoretic perspectives in order to effect the reality that it theorizes (Resnick and Wolff, 1987: 33-37).

[13] This assertion of a dualistic ontology constitutes an attempt to come to terms with the particular relationship of overdeterminist ontology and epistemology.  Overdetermination, in this interpretation, precludes the existence of an outside to material existence (i.e. processes existing outside mutually constitutive relations), implying ontological closure/holism.  The irreducible complexity of totality implies, however, that theoretic processes produce constitutive outsides (Laclau, 1997), through which processual overdeterminants elude theoretic integration.  The anti-essentialism of Marxian ontology arises, thus, from its theorization of a totality characterized by overdetermination and its incapacity to achieve closure in theoretic examination of the processual parts of totality.   

[14] This subsection addresses arguments about class and the city also directly addressed by Tajbakhsh (2001), relative to Harvey, Castells, and Katznelson, in a manner that references the overdetermination of the social identity of individuals (2001: 169).  Acknowledging the relevance of this account, I want to emphasize that my approach and, in particular, my prioritization of class as process and, hence, my conclusions are at variance with Tajbakhsh’s position.