Thursday, September 13, 2012

The territorial logic of the capitalist state


precapitalist territories
  The movements from Tahrir to Liberty Square have in different ways posed the question of space and political authority.

  But perhaps there has been a bias toward approaching this question mainly on the level of international or transnational action, whether it be in the form of ‘globalization’ or imperialism.  Somehow, and I know this bothers you as much as it does me, New Labour’s Regional Spatial Strategies were just not as interesting as the Bush administration’s ‘Green Zone’ strategy in Iraq.  

  It is natural enough that the ‘spatial turn’ of social theory was linked particularly to the question of imperialism in the last decade.  This is not only because the US army corps of engineers could be found building Xanadu compounds, permanent military bases and separation walls in Iraq, which just is an inherently interesting type of spatial activity.  It is because when a state projects political authority outside of its sovereign, bounded territory, and begins the task of organising the political space of another nation, the seeming naturalness and obviousness of the relationship between space and political power is necessarily problematized.  Territorial authority looks very clearly like what it is: an artifice, a production, multi-tiered effort of coercion, culture war, political organization, economic strategy, material incentives, symbolic organization, the construction, reconstruction and moving of internal frontiers, and so on.  Suddenly, the means by which we ourselves are dominated become visible.  Suddenly we get to see how it is all constructed: it isn’t called ‘state-building’ for nothing.  We can learn a lot from the cities defiled by empire, as Derek Gregory has shown us.  

  And, to be fair, things aren’t quite as one-sided as I am painting them.  It is true that the organization of cities, metropolitan boroughs, counties, and regions, was much more of a going concern during the ‘urbanisation’ boom of the Sixties and Seventies.  Henri Lefebvre and his student Manuel Castells were took the lead as social theorists throwing into question the embedded assumptions of Chicago School ‘urban theory’, partially in response to the jacqueries of 1968, and the French state’s urban rationalisation project.  Each understood in different ways – Lefebvre as a marxist-humanist, Castells as an althusserian - that the ideology of ‘urbanism’ was symptomatic, coming just as the urban-rural divide was really dying, the city in its old form as a sort of isolated hive of commerce and culture finished.  This ideology adverted to certain real long-term trends, intimately connected with the state and its functions in allocating production facilities, and ensuring the conditions for collective consumption (of housing, health, education, municipal lighting, rubbish collection, etc.) in its extended sense allowing the reproduction of labour power.  But the explanation of these issues was obscured by the deployment of an object, the ‘urban’, which lacked rigorous definition.  They understood the ‘urban’ to refer to a set of social predicaments, political antagonisms, and so on, among and between definitely social classes and interests, which required a more sophisticated socio-spatial analysis to cope with scalar reorganization that was global and affected not just the metropolitan area, but also the status of national states as the strategic basis for production.  This type of analysis, sometimes leavened by poststructuralist additives, has been sustained through the years by the likes of Neil Brenner, David Harvey, Ira Katznelson and Mike Davis.  Arguably, such work has become more important in recent years.  For example, Harvey’s Right to the City, though simply following up on his long-term research agenda, seems very Zeitgeisty in the wake of Occupy.

  Still, there’s a logic to focusing on the axis of imperialism, since the emphasis on interstate relations and antagonisms forces one to consider the national territory not as a spatial given but as a social relationship.  What I want to do in this, another of my exceptionally long, finger-wagging, listen-up-and-do-as-you’re-told posts, is try to explain something about the capitalist state and its spatial logics.  I start at the inter-state level, by traversing some of the issues thrown up by imperialism, and then drill down to national states and their ordering of space.  Let me start with the ‘new imperialism’.
***
  Theories positing a 'new imperialism' pivot on the intersection of two logics of power, conceptually separate though inextricable in practise. According to David Harvey and Giovanni Arrighi, these logics are: the capitalist logic and the territorial logic. (I’m aware that Alex Callinicos also aligns with this tendency, but I see his dichotomy as being slightly different, focused on the relative autonomy of the geopolitical from the economic logic). Harvey assigns a distinct type of spatial organisation to each of these logics of power. Thus, the territorial logic is that of states, whose power is based in command of a determinate territory and "the capacity to mobilize its human and natural resources". The capitalist logic of power produces capillary flows of capital accumulation which flow "across and through continuous space, towards or away from territorial entities". 

  Lefebvre looms large in much of this thinking, especially for Harvey.  For Lefebvre, the relationship between the state and space has to be constituted along several axes.  First, there is the production of a national space itself, a national territory that bears the marks of human generations, classes, political forces, etc.  Of course, the capitalist state by no means coincides with the nation – but as we will see, this is beside the point.  Capitalist states organise territory as national space.  Second, the state constitutes within the territory a matrix of institutional spaces appropriate for a social division of labour, and the imperatives of political dominance.  Each of these spaces, from the borough to the post office to the police station, condenses a system of social expectations and responses, which become so 'natural' and 'obvious' that they are never articulated.  Third, the state composes a 'mental' or imaginary space, a set of representations through which people live their relationship to the people-nation, the state and the territory.

  According to Lefebvre, the spatiality of the state constantly comes into conflict with "the pre-existent economic space that it encounters", "spontaneous poles of growth, historic towns, commercialized fragments of space that are sold in 'lots'":
"In the chaos of relations among individuals, groups, class fractions, and classes, the state tends to impose a rationality, its own, which has space as its privileged instrument. The economy is thus recast in spatial terms - flows (of energy, raw materials, labor power, finished goods, trade patterns, etc.) and stocks (of gold and capital, investments, machines, technologies, stable clusters of various jobs, etc.). The state tends to control flows and stocks by ensuring their coordination. In the course of a threefold process (growth - i.e., expansion of the productive forces - urbanization, or the formation of massive units of production and consumption; and spatialization), a qualitative leap occurs: the emergence of the state mode of production (SMP) (mode de production etatique). The articulation between the SMP and space is thus crucial. It differs from that between previous modes of production (including capitalism) and their manner of occupying natural space (including modifying it through social practice). Something new appears in civil society and in political society, in production and in state institutions. This must be given a name and conceptualized. We suggest that this rationalization and socialization of society has assumed a specific form, which can be termed: politicization, statism."  (See Lefebvre, 'Space and State', in Neil Brenner and Bob Jessop, State-Space: A Reader, Blackwell, 2002)

  I dare say the concept of the 'state mode of production' is theoretically extravagant, stretching the concept of the mode of production beyond breaking point, but the thrust of this is clear: there are two spatial logics of power, one spontaneous, random, commercial and capitalist, coming 'from below'; the other planned, ordered, rationalised, non-capitalist, coming 'from above'.

  Equally, some of the inspiration for this distinction may have to do with the work of the geographer G William Skinner, who certainly influenced Arrighi, and the sociologist Charles Tilly, who extrapolated from Skinner's conclusions.  In an important book, Coercion, Capital and European States 900 - 1900, Tilly wrote:
"G. William Skinner portrays the social geography of late imperial China as the intersection of two sets of central-place hierarchies ... The first, constructed largely from the bottom up, emerged from exchange; its overlapping units consisted of larger and larger market areas centered on towns and cities of increasing size. The second, imposed mainly from the top down, resulted from imperial control; its nested units comprised a hierarchy of administrative jurisdictions. Down to the level of the hsien, or county, every city had a place in both the commercial and the administrative hierarchy. Below that level, even the mighty Chinese Empire ruled indirectly via its gentry. In the top-down system, we find the spatial logic of coercion. In the bottom-up system, the spatial logic of capital. We have seen two similar hierarchies at work repeatedly in the unequal encounter between European states and cities."

  Again, it seems that the spatial logic of capital is identical with the spatial logic of commerce, a spontaneous system of flows, coming from below, 'bottom-up'; while the spatial logic of the state is rationalising, ordering, coercive, 'top-down'.  Two spatial logics, two types of power, each so distinct that, for Lefebvre at least, the state's type of power even rises to the level of being a mode of production.  

  This has, superficially at least, a topographical similarity to the Deleuze and Guattari couplet, territorialization-deterritorialization.  You have on the one hand the oedipalized territorialities such as the nation, the church, the family etc., imposing themselves hierarchically, top-down.  These are dominant in the feudal era.  And on the other, the "deterritorialized" flows of capital, obeying a logic much like that of free-flowing desire, destroying the Oedipal territorialities.  The capitalist logic arises from the intersection of two developments, two decoded flows - those of labour, in the form of the free worker, and those of production in the form of money-capital.  And so these twin logics of territorializing political authority and deterritorializing capital, though carnally enmeshed*, tug at and militate against one another.  This is a common image in social theory.
***
  My reservation about 'new imperialism' theory was at first an inchoate and very slightly philistine feeling that I simply didn't 'get' the seemingly exaggerated emphasis on space.  Bob Jessop pointed out a decade ago that the turn toward 'globalization' as the master-concept of social theory produced a pronounced 'spatial' turn that was long overdue for reconsideration if not revision.  More specifically, I didn't get the emphasis on the territorial as a drive logically distinct from capitalist power.  It seemed to me that territorial space was just one aspect or moment in the development of concrete social formations in which the mode of production took root, just one of the aspects that the mode of production organised.  Could it really constitute a logic of power distinct from, but equivalent to, capitalist power?

  Robert Brenner's critique of Harvey's 'new imperialism' theory zones in on this problem.  The capitalist logic of power, says Brenner, is clear.  There are a set of imperatives (production for profit, reducing costs, etc.), and a set of  mechanisms (intra-capitalist competition, pricing, etc.) by which those imperatives are enforced.  Moreover, there is a strong empirical basis for its existence.  But the territorial logic of power, the state logic, is not at all clear.  There appears to be only a vague set of determinations behind this logic, no obvious imperatives or raison d'etre driving it, and little in the way of an empirical basis.  Insofar as the state planners have an interest in defending a territory, expanding their extra-territorial dominion, etc., it is not clear that this is distinct from capitalist interests.  In explaining this, Brenner sticks closely to a 'homeostatic' model of the capitalist state.  The state’s functioning and wealth being derived from capitalist growth, its managers will tend to act in accord with capitalist imperatives.  And he suggests that, independently of his conceptual framework, that is actually Harvey's approach in his explanations of concrete imperialist actions.

  Despite my misgivings about the 'new imperialism' theories, I can't assent unreservedly to this critique.  Though Brenner pinpoints what I see as a real problem with the 'new imperialism' theories, his critique is bound up with a set of assumptions I don't share.  Brenner argues that there is 'rational core' to Harvey's theory in that it seeks to explain the apparent disjuncture between what might be in the interests of the capitalist class, and the actions undertaken by the state (wars, etc.).  In place of ‘two logics’, Brenner fulcrums his explanation for this gap on the dysfunctional relationship between a state form (the national state) and capitalist imperatives (which continually transgress national boundaries).  For Brenner, the world system being compose of many sovereign states is an historical fact which emerged in the context of feudalism, not capitalism; its survival is not essential to capitalism.  Capitalism transformed extant territorial states into capitalist states without actually altering the multi-state character of the world system.  This argument is given considerable historical depth and detail in Benno Teschke's The Myth of 1648  A world-state, Brenner argues, would be far more functional for capitalist growth than the national state, since as capital internationalizes, national states cannot implement strategies to resolve international deadlocks in production or make adjustments to overcome global imbalances, with the necessary degree of coordination.  The problem here is that while this might explain the general tendency toward irrational conflict between states, and their failure to coordinate rational policy responses to capitalist crisis, this doesn't actually explain why a capitalist state under a particular management might pursue policies that are seemingly at odds with any attempt at a national level to overcome such imbalances, resolve crises of production and so on.

  I think it would be better to start answering that question with the fact just as there are many states, that there is no single or general capitalist interest, but rather many capitals, nationally constituted and constitutively divided into fractions with a hierarchy of power between them.  The hegemonic fraction is that which leads the others, and dominates the state at any given moment.  And even then, there is no absolute agreement on interests or strategy within a given fraction - hedge funds, for example, may differ from investment banks as to the best way to raid social security.  The more determinations you add, the more complex the divisions become, and their resolution is contingent on political and ideological leadership.  Rather than assuming a homeostatic model of state action, according to which capitalist interests exert a long-range regulatory effect on state power, this approach places the issue firmly back on the terrain of political praxis, with all its implied dysfunctions.  And, in fact, Brenner's explanation for the causes of the Iraq war points to this Gramscian problematic.

one does not simply  I have other issues.  I think the account of why capitalist states are determined by capitalist and not territorial logic, implies an inadequate mechanistic model of causality.  This is a criticism that applies to other political marxists such as Teschke and Lacher, and also in my opinion to theorists such as Fred Block who also deploy a homeostatic model of the state-capital relationship.  Intriguingly,  the same conclusion regarding the contingent nature of the state system has been reached by those deploying an expressive model of causality.  The critical theoretical impulse of the state-derivation school, beginning in West Germany but taking off in the Anglophone left through the work of John Holloway, Werner Bonefeld and others, was to treat the state as a fetishised form of the capital-relation. Influenced by Evgeny Pashukanis, they thought that just as he derived the legal form from the commodity form, so they could homologously derive the state form from the capital form. This being so, the most important thing about the state form was the way in which it condensed at a general level the capital-relation, becoming in a sense the ideal capitalist, in order to remedy the deficiencies and dysfunctions in the circuits of production and exchange.  The capitalist mode of production being world-wide in scale, there was no necessary reason why the states system should take the form of national states. Thus, Claudia von Braunmühl’s essay on the nation-state in Holloway and Picciotto’s seminal State and Capital argues that this pre-structuration of the state system by feudal relations has been imposed on capitalism.  Capitalism’s systemic opportunism allowed it to take hold of the international states system and transform it, but otherwise it is not essential to the system.  There are some extremely interesting aspects of this approach, which I’ll come back to, and the state-derivation approach has produced some rich, suggestive analyses.  But, to paraphrase Bernard Shaw, it is in this respect like a library: excellent to borrow from but otherwise to be avoided at all costs.  

  I am not convinced that the existence of many states under the dominance of the capitalist mode of production is merely an historical legacy contingent to capitalism, which could in principle be superseded.   This is not, I insist, to lapse into that tacit functionalism according to which what exists under capitalism must exist for the good of capitalism.  I can well see that sovereign, territorial states pre-dated capitalism and imposed a structuring effect on the emerging capitalist world order.  I simply do not see that capitalism could provide the material basis for anything other than a multi-state system.  And it seems to me that to say the multi-state system is an historical legacy and nothing more is to explain away rather than explain the problem of the relationship between the multiple scalar and spatial levels of political organisation and the capitalist mode of production.  The development of capitalism seems to entail a particular type of territorialisation of political power, and it's worth trying to understand why.  I think neither the mechanistic nor expressive models of causality can serve us well in explaining this, and I will later indicate the relevance of Althusser’s notion of structural causality.
***

  What does the territorialisation of political power consist of?  Territoriality in its simplest sense is the disposal of a particular bounded space as a base for achieving a particular social outcome.  A church, a police station, a school... all are, among other things, bounded spaces within which a certain type of social organisation is performed; containers for certain sets of social relations.  Territorialising political power, then, is nothing other than a process of binding political authority to a determinate space in which the goals of the politically dominant class, the ruling class, can be achieved.  

  What is distinctive about capitalist state territoriality?  We can start with the problem of sovereignty.  In the international system, the modern state is a sovereign power to the extent that it not only exercises a monopoly of legitimate violence and political authority within its bounded territory, but also demands recognition by other states of its exclusive right within that territory. In exchange, it recognises the same exclusive right of other states.  This notion of sovereignty seems to have origins in feudal property relations, in absolutist states organised on the basis of kingship, patrilineal descent and personal rule. The concept of sovereignty described the absolute right of kings and princes to dispose of their population and territory, to instruct them in their faith, and command them as they saw fit. However, sovereignty takes a particular form under capitalism that is quite distinct from that of feudal states: it is the sovereignty of the people-nation rather than the king as god-incarnate. 

  This redirects the discussion from inter-state relations to social relations in general; the capitalist state is not just a territorial state, but a national territorial state.  It is a national space that the modern state organises, maps, and attempts to bring under a grid of intelligibility.  But as I have also said, the state does not strictly coincide with the nation.  Many states, like the UK, are multinational; and many nationalities have no state. Nor, obviously enough, are the state's powers simply contained within the boundaries over which it exercises sovereignty: one effect of the internationalization of capital is the internationalization of capitalist states. Nonetheless, the capitalist state specifically identifies itself as a national state, and cannot be indifferent to nations. In some cases, it suppresses them. In some cases, it organises them in a multinational territory, reproducing national spaces through its distribution and scalar organization of sites of power as part of the logic of political domination, while elaborating a superordinate nationality as part of its symbolic organisation of the territory it rules. So, nationality cannot escape the logic of statehood, and the territory of the state is always national.

  One strategy for explaining the relationship between nation-state and capital is to relate the development of the national state to the requirements of a unified market. As Fernand Braudel put it:

“A national economy is a political space, transformed by the state as a result of the necessities and innovations of economic life, into a coherent, unified economic space whose combined activities may tend in the same direction. Only England managed this exploit at an early date. In reference to England the term revolution recurs: agricultural, political, financial and industrial revolutions. To this list must be added- giving it whatever name you choose - the revolution that created England's national market.” (Afterthoughts on material civilization and capitalism, 1977)

  This would seem at first glance to explain the relationship between the spread of capitalism and the shift from fragmented polities of various scales and types with nebulous boundaries - city-states, principalities, communes, sprawling patchwork empires etc. - to rationalised, imperfectly homogenised national-states.  Capitalism’s need for a unified space within which production and exchange can be organised gives rise to various ways of organising space – the growth of towns and cities connected by transport and communications (in a single dominant language), unified by a common currency, protected by military installations, within a territory delimited by borders and protected by a sovereign authority.  But returning to Braunmühl’s argument, cited above, we are reminded of a methodological principle derived from Lukacs – the primacy of the totality over individual instances.  It is insufficient to take the nation-state itself as the starting point, the self-sufficient unit from whose aggregation a world-system arises: “An international system is not the sum of many states, but on the contrary the international system consists of many nation states. The world market is not constituted by many national economies concentrated together, rather the world market is organized in the form of many national economies as its integral components.”  

  The theoretical basis for this assertion is the typical state-derivationist approach of attempting to derive the state form from something in the general concept of capital.  In this case, it is at first the world market itself that is implied in the general concept of capital, as capital has an innate tendency to drive through and beyond national boundaries.  Through the mediation of the world market, localised centres of production and accumulation form a totality, and from that arises the material basis for a world state to carry out the tasks of an ideal-capitalist.  The fact that there is not one world state but many states is, then, an historical accident; nonetheless, these states have fallen under the dominance of the world capitalist system, have been transformed into political organisations capable of organising the world market in some way, and thus should not be seen as simply a series of nation-states having a relationship of exteriority to one another.  They are intrinsically linked through the world market upon which they arise.  Clearly, then, if the unification of the market is what is at stake, it is difficult to explain the perpetuation of national states except as a contingent fact.

  However, there is no reason to start by inferring or deriving the capitalist state form from commodity circulation.  We can certainly concede the methodological primacy of the totality.  But we should remember that the capitalism which emerges in Marx’s Capital is a complex totality composed of many determinations.  Each new determination, as Callinicos points out in Imperialism and Political Economy, is connected to the previous in a dependent but non-deductive way.  So Marx moves from the analysis of the commodity form to the struggle over the eight hour day, or financial markets, or another aspect of capitalist relations.  And while each stage of analysis may depend on the foregoing explications, it is not directly implied in them.  Capital markets are not implicit in the analysis of the commodity form; there is no reason why the state form should be implied by the general form of capital either.  That being the case, the capitalist state form must be seen as part of a complex totality of mutually articulated elements, each exerting a reciprocal effectivity on the others, the organization of the whole determining the effectivity of each element.  The capitalist state is overdetermined by the organization of the complex totality of which it forms a part.

***

  From this perspective alone can one start to do justice to the relationship between the national state and capitalism. The rational kernel of the state-derivation approach is that capitalism doesn’t merely impose a new organization on territorial entities which otherwise remain essentially the same, but rather take hold of an alter their materiality.  This is true of all the spatial units that occur under capitalism. Castells denounced ‘urban theory’ for presuming that the urban retained the same content, the same meaning, through centuries of change, across modes of production and the stages in their development.  Urban space meant something quite different in medieval Europe; towns and cities represented a different spatial organisation of the social division of labour.

  In Marx’s terms, the social division of labour is distinct from the technical division of labour, in that it arises from social functions related to class.  Under capitalism, the social division of labour is organised around personalized bonds between the feudal lord and the peasant or serf.  The major form of extraction was directly political – the ruling class took tribute by means of coercion – and this tended to produce both a parcellisation of political authority and disarticulation of space.  The rural-urban divide, the patchwork of spaces ranging from estates and small towns to great commercial centres and ports to walled cities, condensed a particular social division of labour.  In the countryside, the producer was bound to the soil working for nobles of various rank on estates of varying size with porous frontiers and no clear boundaries.  Towns varied greatly in size and function; some were enclaves of relative commercial freedom, particularly port towns connected to world markets; smaller towns were usually abutments to large feudal estates.  Their inhabitants did not live off the land and produce tribute, but rather lived off petty commodity manufacture, or various types of trade serving the luxury consumption of the feudal ruling class.  (On the rural-urban divide in the middle ages, see Rodney Hilton).  The major form of ideological dominance being religious, there were overlapping frameworks of sacred and secular power, the church acting as landlord, tithe collector, and symbolic guarantor of the unity of the state as the body of Christ.  This specific articulation of economic, political and ideological power was also the basis for the checkered system of miscellaneous polities, from communes  to city-states to empires.  So, the territorialization of political power under feudalism, based on bonds to lord and land, was such that spaces tended to be irregular, reversible, turned in on themselves (though bonds to lord and land) yet simultaneously open (through extensive migration across intersecting boundaries).

  Where the capitalist mode of production took root, however, producers obtained their famous dual freedom, from both lord (bondage) and land (the means of labour); they were drawn into relations of production mediated through exchange, selling their labour power to capitalists who procured it as just one element in a productive process intended to produce a profit.  However, as Poulantzas pointed out, this did not entail deterritorialization; such a schema relied on a naturalist image in which territory was assumed to have a continuous meaning, connotatively linked to ‘rootedness’ in determinate plots of land.  Rather, the capitalist division of labour entailed a different type of territorialization.  Production, circulation and exchange now demanded a spatial matrix of imperfectly homogenized sites, segments of space each carefully delimited by clear frontier marking insides and outsides and linked to a social division of labour – factories, hubs, supply chains, shopping centres, terraces, conurbations, condominiums, and so on.  So while the movements of money, capital and labour would tend to push beyond these spaces, they must cross frontiers in order to do so.  This system of boundaries is necessary to organise the labour force, the distribution and storage of goods, communications, transport, consumption, residences, and so on.  It is necessary to help regularise an already anarchic system of production and minimise its dysfunctions: for example, they help impose a general sedentarization on the labour force (see James Scott on this) that makes its supply more predictable and its constituents intelligible.

  The specific combination of cooperative and competitive relations in the division of labour also has effects on the spatial matrix.  Production, distribution and exchange must necessarily take place in a cooperative manner, meaning that capital units are locked in a relation of interdependence.  This will produce a tendency toward clustering, as functionally associated capitals reduce their distance from one another: it makes sense, for instance, that large manufacturing enterprises would tend to cluster in industrial estates near large workforces with access to main road; or that commercial enterprises would cluster on high streets in pedestrian and motorist accessible centres where consumption can take place.  On the other hand, this cooperative effort is structured by competitive accumulation.  Some capitals will succeed better than others, and over the long-term there will be a concentration and centralization of capitals, which themselves attract chains of supporting industries, producing spaces (towns, cities, even countries) which work as privileged centres of productive capital, and by extension other spaces that are underdeveloped and neglected.

  Political authority under capitalism, rather than being directly embedded in those sites through a chain of significations linking land to labourer to lord, acquired a formal separation or relative autonomy from them.  Indeed, part of its role was to help constitute this new spatial matrix by standing in a formal sense ‘outside’ it, while ‘intervening’ constantly.  The scare quotes are necessary, because it is clear that in no real sense does the state have an external relationship to the spheres of production or exchange.  This is where the state-derivation approach produces an important insight: breaking with the fetishised notion of the state, with the legal, constitutional image of the state as simply an external guardian of civil society, it treats the state as a social relationship, actively involved in the constitution of the totality of social relationships in part by separating off aspects of them and deeming them ‘political’ as opposed to ‘economic’.  This is consistent with Corrigan and Sayer’s important argument that ‘the state’ as such is a fiction, a ‘mythicized abstraction’; it is through the state relation itself that the social categories are produced to give it its seeming legal and institutional determinacy, its “misplaced concreteness”.

  Still, despite the above, and despite the spatial metaphor deployed, this ‘standing outside’ adverts to a real political relationship which is the state’s relative autonomy from social classes.  As Claus Offe put it, this relative autonomy is necessary to capitalism because only a “fully harmonious economic system that did not trigger self-destructive processes of socialization could tolerate the complete positive subordination of the normative-ideological and political systems to itself."  It is the fact that capitalist production is not a self-sufficient system, that it has inherent crisis tendencies, and arguably the fact that is articulated with other systems (ecological, biological, etc.), which makes it so inherently unstable and requires a state with the freedom to provide a spatio-temporal fix.  Another relevant feature of capitalist production is the ‘isolation effect’ it produces in social classes.  Because it is a system of competitive accumulation among many producers, and because capital is constitutively divided into fractions, the capitalist class finds it impossible to constitute its political dominance over the popular classes without the state, which cannot therefore be an ‘instrument’ or ‘tool’ for the capitalist class as such.   So the apparent extrusion of political authority from the organization of the spatial matrices of production, circulation and consumption is actually nothing other than the formal separation of the political from the economic; the state remains deeply involved in and articulated with the processes of capital accumulation, constituting the segments of space through its schools, police, armed forces, councils, parking authorities, free enterprise zones, etc.  And through its action it seeks to unify and homogenize those spaces; but how?

  In the capitalist mode of production, the dominant form of ideology is no longer religious but political; in normal circumstances, the capitalist state presents itself as a popular, representative state (even if not actually democratic).  It does so firstly by binding itself to a nation, an ‘imagined community’, usually with a shared language.  But to represent the nation as such, it must dissolve classes at an ideological level into individualised subjects, who are then cemented together through the state; the dominant ideological form this takes is legal; the law produces the ‘free and equal’ subject of the bourgeois nation-state.  This is connected to the enclosure of a ‘national’ space, which is obviously by no means a natural space (though of course national expansiveness is necessarily responsive to natural resources, and the spatial matrix of production is warped around them).  Just as the segments of space at the level of factories, bureaucratic offices or towns are circumscribed by a clear frontier as part of the logic of organising the social division of labour, so the state constitutes the national space by erecting a frontier around it, a system of exclusion and filtered admission (of labour, goods, etc) which is operated on behalf of the nation.  

  Not only that, but the state effectively operates a system of internal borders, whereby those who are deemed non-national or anti-national can be confined, brutalised, hyper-exploited, etc. – this can range from detention centres for asylum seekers to concentration camps; from Jim Crow laws restricting movement to secret prisons.  This too has a certain relationship to the social division of labour, insofar as the latter is constituted by politics and ideology.  Capitalism has always shown a marked tendency to stratify labour forces according to principles of race, nationality, religion, gender, ethnicity, and so on, in a way that enhances the political dominance of capital over labour and increases the rate of exploitation of all workers over the long-term.  As Roediger and Esch’s accounts of ‘race management’ would demonstrate, this is not something that simply takes place at the level of the state; such strategies are implemented and experimented with directly in productive enterprises.  But the state also develops strategies for the control of labour forces, for example by obstructing the mobility of some workers to discourage migration at some points or render migrant workers insecure at others, or implementing material incentives in a gendered way so as to preserve a family structure in which women perform the labour of reproducing labour (ie maintaining a household, raising children, feeding male workers etc).  The system of both internal and external frontiers is part of the organization and disciplining of the pyramid.

  This directs one’s attention to what the legal concept of the border, as simply an arbitrary political cleavage separating nation from non-nation, obscures: the fact that the frontier is a set of social (economic, political, ideological) relations, mediated through the state, between the contending classes bound by it; between the many capitals based within it and those beyond it; between national oppressed and dominant groups, and those beyond the nation; and between social formations unified by respective national states, whether imperialist or non-imperialist.  The transgression of frontiers naturally also represents one moment in a given social relationship, be it oppression (refugee flows), exploitation (labour migration), social resistance and class struggle (breaking out or breaking in), or imperialism (invasion, bombing).  The point I’m making here is that the most important fact about national states is not that there are many of them, although logically there must be and this is important; it is the social relations embedded in them, which make them national states.

***

Tahrir  So, rather than deriving the appropriate type of territoriality of the capitalist state from one of the general formal elements of the capitalist mode of production, I have suggested multiple levels of determination beginning with the overall social division of labour, proceeding to the organisation of political authority specific to capitalism, the dominant form of ideological domination under capitalism, the type of relation between capitals, and between different groups of workers, and the effects of these determinants on each other.  I am not claiming to have been exhaustive, but approaching it in this way gives one a superior perspective on the different orders of scale and space that characterise the capitalist world system and particularly allows one to understand why the key strategic base at which territorial statehood is organised is the nation and is likely to continue to be the nation.  Of course, the specific unity between nation and state is a practical unity, not one that is given in the theoretical understanding of capitalism.  Nations are not essential to capitalism.  However, in any counterfactual scenario of capitalism’s emergence and development, the world system would most likely be constituted by a graduated system of spaces at the strategic centre of which would be something like national states.

  And it’s worth thinking about how this hierarchy of oedipal territories, this very tightly controlled grid of social spaces from the main public square to the prison, dense with symbols and ideology as they are, are so important to capitalist political dominance.  I don’t claim that actually controlling this or that bit of space is the most important thing about capitalist territoriality – no, it is the social relations embodied in the space that is central.  I have said these segmented spaces are ‘containers’ for social relations in which an implied set of social expectations and responses are condensed.  Because of this, there is a strategic hierarchy; some spaces are worth more than others, because the social relations they embody are more important than others to reproducing the system.  

  A certain approach to anticapitalist struggle focuses on taking over spaces, and creating autonomous zones which are simultaneously protest, pedagogy and prefiguration.  This has been tried with the Occupy movement, and has demonstrated some real strengths in all three capacities.  But it has clearly reached its limits inasmuch as the capitalist state has learned, through trial and error, how to regain control of such spaces.  And to the extent that the strategy was bound up with a longer term perspective of, piece-by-piece, liberating space and building communism on a cellular level, it was never going to work.  For, if it is worth claiming autonomous spaces, then it has to be asked why those actually claimed outside of actual revolutionary situations (and then only for a short period) have always been strategically negligible, of marginal importance to the reproduction of the system.  The answer is most likely that the politically important spaces are always very well organised and manned, while it would be unwise to occupy strategically important productive spaces unless the workers already agree with you, in which case they should be doing it.  That leaves public and semi-public spaces, and a few vacated buildings.  Not exactly the Paris Commune; not even the parish commune.  And since the national state is the privileged level of the political organization of the territory, capturing visible but strategically marginal space is always at best a short-term tactic, doomed as it is to encirclement and shut-down in short order.  The only sensible answer is to re-focus on the social relations that are constituted through a particular organization of space, and try to organize the agencies best placed to disrupt their reproduction.

*That means exactly what you think it does.