The Future of Insurrection

Section 1: The composition of insurrection

What is insurrection?

'The goal of any insurrection is to become irreversible. It becomes irreversible when you've defeated both authority and the need for authority' (CI)

'The catastrophe is not coming, it is here. We are already situated within the collapse of a civilization. It is within this reality that we must choose sides. To no longer wait is... to enter into the logic of insurrection' (CI)

'It's enough just to say what is before our eyes and not to shrink from the conclusions' (CI)

What strategies and orientations can develop insurrectionary anti-politics into a movement actually able to destroy global capitalism? This is the question taken up by The Coming Insurrection, as well as by author such as Bonanno. I aim here to use insights from The Coming Insurrection to open onto discussions of various aspects of the future of insurrection. The purpose will be to think through strategic implications of attempting to use a mainly expressive form of action for strategic purposes, and ways to deal with the obstacles faced in the process.

But first of all, what is insurrection?

Insurrection differs from revolution in being an attack on the existence of state power, rather than a seizure of such power. It follows in the tradition of Walter Benjamin's idea of 'law-destroying violence', which is directed against the capability for use violence to make or preserve laws. It is not instrumental violence to subordinate others, but rather, exists beyond the mythology of statist violence, destroying the power of death for the sake of the living. Bonanno's theory of insurrection relies on a concept of social war, which refers to an irreducible antagonism between included and excluded. Insurrection for Bonanno involves the rejection of alienation, especially of subordination to production, and involves both an affirmation of life and desire and assault on the structures of power. 'Unfortunately civil war is an obligatory road which must be passed in any historical moment of profound, radical transformation'. Yet it must also be playful, generating excitement and a sense of empowerment against the social system as death-machine. Insurrection pits active force against reactive force, and is the point of explosion of accumulated discontent.

It is fundamentally connected to non-renunciation, the refusal to compromise on desire. It is thus connected to an active, affirmative type of desire. It is also fundamentally connected to affinity-networks and bands, as opposed to organisations. It most often arises from standpoints of exclusion or marginality, as opposed to those which are included but exploited. Insurrection has at least three political components. It has an affective or expressive component: attacks which respond to indignity and violation, which are psychologically liberating and inspiring. It has a strategic component: it imposes costs on repressive or oppressive forces, and serves to carve out spaces of autonomy by altering the balance of forces. And it has a prefigurative component, with each act of insurrection pointing towards and attempting to produce in the present a total insurrection resulting in the destruction of the system. Insurrectional agency is effective when the three components are articulated. And this creates issues of their interconnection and the passage between them: how insurrectional acts which are affectively motivated and expressive can nevertheless serve instrumental purposes such as carving out spaces of autonomy and imposing costs, and how and when these spaces and costs reach the point of bringing down the system. We can think of issues ranging from summit protests to squat defence to the SHAC model as examples of how the first connection comes about. The second is more tentative, but raises questions of why for instance the Greek insurrection fizzled out after three weeks. Is there a time-limit on insurrections in the global North, and how can it be overcome?

Just-in-time repression

'The lid on the social kettle is shut triple-tight, and the pressure inside continues to build. From out of Argentina, the specter of Que Se Vayan Todos is beginning to seriously haunt the ruling class' (CI)

There is no question that insurrection is growing. This is because the paths of reform and revolution are failing. At present states are becoming less attuned to social struggles, because they are seeking comparative advantage to attract global capital. They are increasingly reluctant to make the concessions they would once have made, to keep social peace. They will accept immovable objects (the Peruvian Amazon struggle for instance) – but only when they are absolutely forced to; and one can normally expect all viable tactics of repression to be employed first. The field of insurrection thus comes to overlap with the fields of reform and revolution, which can succeed only by way of insurrection (though the revolutionaries and reformists are slow to learn this).

Things have changed. Gramsci's old notion that the frontline of capitalism is now buttressed by the 'trenches and fieldworks' of civil society is no longer valid. It spoke of a Fordist and corporatist era which has passed. Today, capitalism is once more engaged in a war of movement. More than this: it is like an army which has all its troops on the frontline. It has corroded all its deep supports, such as legal due process and civil rights, as too costly to maintain. It has pushed its forces of repression further forward, onto a frontline where people can barely speak out without facing repression. But behind this frontline there is an open field all the way into the system's territory. The mentality of just in time production has been expanded into the fields of politics, security, repression. Just-in-time security means events like 911 are always just about to happen, only just averted – in Powell's words, they had the information it would happen, but there was too much information to handle, to filter.

So, too, the field of protest: the frontline forces are vicious but are always only just in time to prevent an event. Unexpected events, like the unrest on the first of Britain's student protests, the Melbourne taxi drivers' protest which shut down the city, the 'Anonymous' DDoS attacks on Wikileaks' persecutors, the flash mobs which periodically hit major cities, can flare up out of nowhere, taking the state completely by surprise. Emerging like a snake from the spaces of quiet suffering, they pose a constant threat of 'unknown unknowns' the system cannot handle. The closer insurrection is to these unpredictable modalities of protest, the less it can be pre-empted, and the more the vast space behind the front lines is open to it.

This provides opportunities for exciting events. But there is also a certain danger in the upturn. Anarchists are not affected by who's in power, but the wider field of potential resisters clearly are, and this in turn affects things like the numbers resisting and the general level of energy. This ultimately seems to affect anarchists too. Take the situation in Britain. Today, there is a wave of militant resistance to the new Tory government's cuts programme, with impressive actions in London. Yet it is strange that this has come only now. Things were no better under Blair. Then, a militant opposition emerged (in events like J18 and Mayday), precisely because the broad constituency of resistance was decomposed by Blairism's clever use of incorporation and lesser-evilism. Now, because of a recession and a Tory government, the masses are moving once again. But the future does not lie with those who will be quieted by a change in government. The future lies with those who do not compromise – which is to say, with the network of bands. The danger we face today is the reabsorption of the bands into a movement of resistance hegemonised by the mass. The opportunity is that people can be drawn from the mass into the network of bands by the experience of struggle.

Texts like The Coming Insurrection are charismatic. They resonate on the basis that they make claims which appeal to the reader, subjective truths which are otherwise hidden. Insurrections, too, are charismatic. And it is impossible to know in advance how resonant they will be, given their exclusion from public discourse by the dominant system. Resonance with the hidden transcript, or with psychologically repressed material, or with groups denied a voice, is hard to predict. Insurrection is also expressive. 'Instrumental action relates to only one sphere of the lifeworld, another sphere being... [the] expressive'... [T]he pursuit of expressive authenticity is a form of protest against disenchantment, which is brought about by the rationalization of the lifeworld' (Routledge and Simons 476). The replacement of instrumental with expressive orientations should be one of an insurrection's goals. There is something inherently appealing in meeting state violence with a counter-attack, something which is missing in other responses, for all their usefulness and bravery. 'It is high time for them to understand that we refuse to put up with this abuse any longer' (Black Block Papers, p. 80).

The experience of the excluded and the right to be angry

'We can no longer even see how an insurrection might begin... our sense of the war in progress [is dulled]... We need to start by recovering this perception' (CI)

'No one can honestly deny... this was an assault that made no demands... and it had nothing to do with politics' (CI)

The intensity of experience sustains insurrection beyond its specific goals. Above all, insurrection is a question of intensity. In bourgeois rhetoric, intensity and violence fuse into one another. Intensity is frightening to the system because it does not take part in the politics of inclusion, it does not sell itself to power. The images of the “violence” of insurrection thus fuse real attacks with imaginary violence, with the state's fear of its own collapse.

There are of course dangers of insurrection slipping into roles and reproducing the system's violence, but these dangers are overplayed by critics. Insurrection is not at all a masculine thing, a performance of social roles – it is all about the right to be angry. Similarly, activism of whatever kind if not above all a publicity stunt, not a performance for the mass or state gaze, but something else, an expressive action, an act against or in radical antagonism with the state, imposing costs on it. Fighting the police as enemies stands in the tradition of indigenous warfare, of "popular defense" in Virilio's sense, not the modern warfare which exterminates the enemy as irreducibly evil or which closes space to prevent action.

Another criticism we can safely ignore is the leftist objection that insurrection is an action of a minority, and that images of insurrection are alienating to the majority. Insurrection is performed to bring about a better world, it is not performed for the gaze of the Other. There is no reason insurrectionists need to be accountable to the majority; we are the excluded, those who are not part of the community, so the majority is not part of the same collective as we are. Why should the excluded always be the ones expected to dialogue, compromise, appeal to others? The system has made clear it has no time for such things. It is putting itself further and further from any possibility of dialogue. Leftists tend to assume that capitalist power is nothing but the alienation of our own power. This is true, if the “we” is cast broadly enough (it is alienated life), but it is not maintained by the insurrectionists, the people who resist; it is maintained by others, whose positions are incommensurable with ours. They are not simply seduced by false consciousness or forced to alienate their labour; they actually desire the present system. Hence, we should not imagine that all of this will dissolve in the event that individualism is replaced by collectivity and struggle. For one thing, insurrectional struggle is on a certain level very much individual.

But more crucially, the structural analysis underpinning this view is flawed. Leftism makes excuses for people's reactionary ideologies by taking a starting point of ideological submersion as axiomatic, and imagining community or struggle to be a messianic antidote; if anarchists criticise people for being reactionary, they're prone to call us 'moralistic' (meaning we have our own ethical principles, instead of a historical teleology). Of course, insurrection often transforms those who are a part of it, and many people go through moments of revelation in the face of police brutality. Yet insurrectional bands are most often formed from prior individual refusals; the refusal constitutes the community, it does not result from it. On the contrary, movements which start out as reformist or reactionary do not miraculously become insurrectionary simply because people come together. This is because the real basis of revolt resides in desire, not community. People are not simply products of ideology and subjectification until they miraculously break its spell in revolt; the system needs hooks in desire to draw people in. People vary in the degree to which their desires resist this process of attraction. The facts that the social relations have to be continually reproduced, that fetishism is incomplete and can break down at any point, that systems left to their own devices go into entropy, do not at all affect the fact that the system will not collapse while those who desire it exert dominance over those who do not.

The included (including people who are exploited, but nevertheless identify with the system) have betrayed insurrection time and again. In doing so, they have harmed their own position as well as ours. If the included do not care enough about the most basic rights and needs of the excluded, or do not have the power to force the system to concede - why on earth should the excluded hold back, out of concern for their approval? In any case, the mass doesn't think, it just reacts. Greek “public opinion” supported unrest while it happened, then fell on board with the government line. British “public opinion” was massively against Thatcher at the height of crises, only to return to her at elections. We need a more constant compass than this. The Coming Insurrection is right: there is no prior community, no “we”, only the affinity of those who are linked in aspirations and actions. The ground of resistance is not the community or the majority, but each person's right to be angry and to resist, based on our difference, our refusal, and our non-renunciation. We shouldn't feel a need to form links with others who have no desire to form links with us. Against alienation, yes: alienation from ourselves, from others who resist, from the environment... but separating ourselves from reactive force and those who bear it is not alienation, it is autonomy. In Clastres' account, indigenous groups maintain autonomy by separating not only from the state form but even from one another, to defend the autonomy of each group. We need to get past the simplistic association of separation with alienation. Of course we can, and should, ask how, if at all, we can bring over some of the people who aren't resisting to our side. But we must not subordinate our will to theirs, nor imagine we're doing good by indulging their self-limiting aspirations or their reactionary beliefs, in other words, by compromising on our own desires. We aren't all in it together

Short of the final collapse of state and capitalist power, the maximum effect of insurrectionary actions occurs when it is nevertheless pushed back, rendering the effects of such actions cumulative, and expanding open or liberated spaces. These strategic effects are difficult to produce, and can only come from a fusion of the brain with the heart: using expressive actions to produce instrumental improvements, which in turn reinforce the expressive states productive of further action. Strategy comes easy to organised movements because they can turn mobilisations on and off. It is harder for expressive movements and bands, but it is still possible, because expressive affects have triggers and varying intensities.

Insurrection rejects the state's claimed monopoly on force, largely because it pursues a diffusion of all forms of social power. The mainstream attitude to violence in the global North is like the Victorian attitude to sex. There is at once an emotionally invested prohibition, replete with condemnations and silencings, and an attempt to restrict its legitimate expression to a confined proper context in which enjoyment and excess are excluded... and a proliferation in practice, from structural violence to police brutality to organised crime. The ridiculous outrage about insurrectional violence is not only hypocritical – it isn't really about violence at all. It's about what are known as 'feeling rules', in particular, a prohibition on feeling angry against the system.

A qualification, however. There is a lot of talk in The Coming Insurrection about events, decisions, subjective truths and so on. This is a big theme in French philosophy today, probably lifted from the work of the Maoist Badiou, and can be traced back to the debate between the fascist Schmitt and critical theorist Walter Benjamin. The problem with the use of decisionism in insurrectionism is that the act of decision, the 'sovereign decision' as Agamben calls it, is constitutive of the state. Hence why Schmitt, the founder of this concept, could still be a fascist. We need to be clear on how our anti-politics is different from that of the statist suspension of the ethical, the sovereign decision. For Agamben, this distance is created by being all-inclusive and immanent, which is to say, it rejects normativity, it diffuses ethics. For Virno, by a kind of decision which is not sovereign, but which simply emerges from a distributed network. For Benjamin, in the difference between a violence which founds law and a violence which destroys law, which is to say, which diffuses power. I think, too, that the Invisible Committee's version is different: we do not make the decision, it takes hold of us. It 'will occur to us rather than being made by us'. What these views have in common is diffusion, the replacement of concentrated relations with diffuse relations (ethics for Agamben, process for Virno, violence for Benjamin). Decisionism is associated with the concentration of power, and hence is part of statist reason. We need instead a diffusion of the power to decide.

Insurrection and band societies

'the decomposition of all social forms is a blessing. It is for us the ideal condition for a wild, massive experimentation with new arrangements, new fidelities' (CI)

'We're setting out from a point of extreme isolation... FIND EACH OTHER. Attach yourself to what you feel to be true. Begin there' (CI)

'All affinity is affinity within a common truth' (CI)

There is a special kind of group which is the agent of insurrection. I have variously seen it called a band, neo-tribe, neo-sect, bund, pack, fused-group, subject-group. Anarchist groups are, at least partially, band societies. For some, in all or most aspects of life; for others, as the most emotionally intense aspect of lives also lived less intensely in other political forms. As Virilio observes (Speed and Politics, p. 4), street insurrection reproduces the raiding party of our ancestors. Anthropologists such as Ingold have shown that bands are fundamentally different from societies in the usual sense. The band involves a way of constructing social relations which does without the usual hierarchical props. It is absolutely immanent to everyday life. Militant resistance gets its power from its articulation in everyday life, not only in the moments of insurrection themselves, but in the full set of autonomous practices of which they are a part. (What is sometimes attacked as 'lifestyle activism' is actually the embodiment of this immanence). The band as a social form seems to reappear wherever organisation isn't fully implanted. Bands seem to come naturally to children. Anarchists, and far-leftists too, usually end up in bands, even when they want organisations (like Makhno for instance).

Anarchist bands are somewhat distinct from indigenous band societies, being rather closer to the 'bund'. Schmalenbach's account of the sect or 'bund' suggests that it achieves a social form irreducible to community and society, held together by a special kind of emotional bond he terms 'communion'. Immediate emotional experiences (often produced through ritual) hold together such groups without the mediation of abstract identities or organisations. Existing without a basis in ascriptive ties such as kinship, the bund cannot exist independently from the social acts which constitute it. It must constantly be re-enacted, or disappear. It is thus an absolutely immanent form of social life. Acts of insurrection constitute insurrectional bands, playing the function of ritual. Bands are counterposed to rational, linear history (since are based in immediacy of affective fusion, not representation).

Perhaps the most commented-on discussion in The Coming Insurrection is the critique of milieus and the argument for communes. Milieus are constituted by the problematic aspects of many activist groups, such as informal hierarchy. They are deemed reactionary because they betray truths and are only concerned with their own 'sad comfort'. In contrast, communes arise when people find each other and forma common path. A multitude of communes could replace all the various institutions of the dominant system, forming an entire counter-society. 'The commune is the basic unit of partisan reality. An insurrectional surge may be nothing more than a multiplication of communes, their coming into contact and forming of ties'.

This raises the question of the transition between the two. When do milieus turn into affinity-groups and vice-versa? The difference seems to be defined in terms of their animating social logic and emotional formation. Crucially, communes are defined by 'the spirit that animates them', 'the density of the ties at their core', and not by 'what's inside and what's outside them'. This defines them in distinction both from right-wing networks and from states and other hierarchies. It also establishes them as very close to the categories of band, pack, neo-sect, fused group, subject-group, and bund. A 'truth' is here associated with the intense, immediate emotional connection at the heart of these kinds of groups. The loss of this leads to a 'milieu' as direct connection is replaced by some kind of normativity.

In other words, insurrectionists, communes, are always bands. But not all bands are insurrectionist. Formally, all band societies are rather similar. But they differ in how their identity is constructed. The autonomous kind of band should be distinguished from those types of 'sect' and 'neo-sect' which claim to be the one true way, viewing themselves from the start as something akin to a universal church. Band societies can be reproduced only if they coexist with other bands in a terrain of multiple voices and horizontal connections.

Networks and the everyday

'The maintenance of the self in a permanent state of deterioration, in a chronic state of near-collapse, is the best-kept secret of the present order of things' (CI)

'We count on making that which is unconditional in relationships the armor of a political solidarity as impenetrable to state interference as a gypsy camp' (CI)

If the band is the basic unit of insurrection, networks are necessary to reproduce it across time and space. In The Coming Insurrection, it is argued that we can no longer find each other in sites such as the factory; instead, affinity is formed through everyday insubordination. It has long been argued that coming-together as affinity-groups is already an act of insurrection; Hakim Bey, for instance, views it as defiance of the capitalist distribution of time. The Coming Insurrection stands in this tradition of emphasising affinity.

Major insurrectional events involve bands, but also swarms. The band coexists with the swarm; swarms emerge when bands come together for an event. Movements over time switch between swarms and bands. When swarms decompose, bands tend to come to the fore. A swarm may arise when a number of bands coalesce. The transition between the two requires a degree of critical literacy, dialogue, inclusiveness, avoidance of silencing. Inclusive networks are the means to make bands into swarms and swarms into bands.

If everyday life forms a site in which insurrection can be built, it follows that insurrection is not limited to those acts the system demonises as 'violent'; it also encompasses an entire range of 'nonviolent' approaches: building links among excluded groups and bands, reconstructing subsistence economies, 'social weaving', emotional healing, forming bands and networks which create their own values, the construction of autonomous spaces. The strategies proposed by authors such as Colin Ward and Hakim Bey, or autonomist strategies of 'exodus', of defecting from capitalism and withdrawing life-energies and creativity from it, are not counterposed to those of insurrection, but operate as its everyday level, its condition for reproduction. In practice, insurrections emerge from, and extend, networks of power and meaning already operating in everyday life, often submerged or hidden. (This also suggests that insurrection is in continuity with, not entirely separate from, resistance). It is easy enough to find useful things to do, other than actually fighting the system on the frontlines. But this cannot be a substitute for insurrection. Ultimately, the state will respond with violence to the recomposition of forces it cannot control, and the recomposed bands will either have to deal with dispossession or fight – and defeat – the state. We need to radicalise the idea of diversity of tactics as it applies to protests, embracing interdependency and the insufficiency of each actor to the total struggle as part of a broader radicalisation of interpersonal relations. Not everyone can fight the police; not everyone can forego fighting the police.

Insurrection should thus be part of a broader process of reclaiming life from capitalism and the state. This is not to say, however, that unrest short of a final destruction of the system is unnecessary or unproductive. Small, apparently ineffectual insurrections, often deemed ritual protests by researchers, become crucial means for building the subjectivities, repertoires of action and 'action spaces' which prepare for insurrections which can bring down the system. An event like the Greek insurrection of 2008 is made possible by the more ritualised showdowns of the November anniversaries and other events all year long.

Place

'to keep the riots going for a month, while keeping the police in check – to do that you have to know how to organize, you have to establish complicities, you have to know the terrain perfectly' (CI)

The relation to territory also changes: instead of possessing territory as in state projects, insurrection increases the density, circulation and solidarities of communes, rendering the territory 'unreadable, opaque to all authority'. This requires a proliferation of existential territories: 'the more territories there are superimposed on a given zone, the more circulation there is between them, the harder it will be for power to get a handle on them... Local self-organization superimposes its own geography over the state cartography, scrambling and blurring it: it produces its own secession'. The text portrays this, not as a return to local slowness against state speed, but a surreptitious overtaking of the state. Territory should here be understood in relation to the distinction in geography between places, which are sites of meaning for participants, and space. Capitalism is premised on spaces which are not places, 'non-places' such as airports, hotels and supermarkets which resist being turned into local places.

The reconstruction of local space creates which is sometimes termed 'homeplace', a type of place in which people feel emotionally secure and at ease. The imposition of non-place also imposes generalised insecurity and anxiety. Place, or existential territory, exists in the dense indigenous relations to particular local ecosystems, the detailed spatial knowledge and sense of belonging to a locality of inner-city and banlieue rebels, even (in a mediated way) in the worker's relationship to the factory. The current phase of capitalism (and not necessarily earlier phases) seeks to replace the experience of place with a mixture of 'telepresence' (virtual images) and non-places in which people are controlled and equivalent. In this phase, the restoration of place can be a means to restore autonomy which has been lost by localities. Of course, caution is needed here to distinguish the autonomous construction of place from exclusionary and oppressive types of local identity (such as nationalism and racism), and from a purely defensive orientation to place (such as rural conservatism and working-class nostalgia). An autonomous relationship to space is a localising relationship but also an immanent relationship counterposed to the transcendence of any particular spatial imagining. It is, in the Zapatista slogan, 'a world where all worlds fit'.

Section 2: The power of insurrection

'to know that a certain coexistence will end soon, that a decision is near' (CI)

'any loss of control would be preferable to all the crisis management scenarios they envision' (CI)

Insurrection has power when acts are available to insurrectionists which are not available to hierarchical power. The state tries to destroy such advantages, both through recuperation, bringing in approaches which begin outside, and through repressive countermeasures. Insurrections often arise in a cyclical way. The emergence of a new tactic to which the system cannot respond generates new forms of insurrectional power. These new tactics create cracks in the dominant system, which attract other people amd groups who were formerly disempowered. The tactics reproduce virally. In contrast, downturns in militant activity occur when existing forms of action seem to have stopped producing powerful effects.

The motor of change is the instability of the existing order's ability to 'govern' or 'command'. This relies not primarily on its ability to suppress, but on the persistence of obedience. Each insurrection disrupts or destroys the system's ability to command. Each time, the system will either collapse or recompose. So far, it has recomposed. Of course, neither insurrectionists nor statists can foresee the other's ability to invent new tactics or weapons. So both new insurrections and new recompositions of the system are unpredictable. A final collapse of the dominant system will occur when the system cannot invent new responses or weapons quickly enough to contain the ability of an insurrection to undermine command. This also leaves the question of how to reconstruct spaces outside command once the system has collapsed, or while it is collapsing.

Asymmetrical conflict

'The police are not invincible in the streets, they simply have the means to organize, train, and continually test new weapons. Our weapons, on the other hand, are always rudimentary, cobbled-together... [and] don't have a hope of rivaling theirs in firepower, but can be used to hold them at a distance, redirect attention, exert psychological pressure or force passage and gain ground by surprise' (CI)

The tactics and techniques which form weapons of insurrection and repression, as well as the literal weapons, are constantly innovated on both sides. The Coming Insurrection plays up the state's research capacity to generate new weapons. This is, indeed, a serious problem, though the state tends to develop new weapons modelled on old ones, new weapons which do the same things in slightly different ways (is there really a world of difference between microwave beams, LRADs, water cannons, tear gas and shooting in the air?) In contrast, it suggests the improvised weapons of insurrectionists are necessarily inferior, and implies they do not develop. This is not necessarily true. Firstly, activists through time have innovated a whole range of tactics which later catch on, such as the various innovations in lock-ons, tree-sits and tunnelling in the 1980s. Secondly, there are a great many actors – from smaller state powers to organised crime networks and armed opposition groups – doing research into undermining asymmetrical power. These actors often discover things that are later used in insurrections. Molotov cocktails were invented by the Soviets as a cheap way to fight an invading army. The Internet was originally invented by, of all people, the US military, as a defensive measure against massive assault, before being taken up by hacktivists (consider the Operation Payback actions for instance). Thirdly, age-old knowledge can be rediscovered, as when activists borrow consensus decision-making from indigenous groups. We should look for new vulnerabilities, and tools which exploit such vulnerabilities. In China, the next big wave of asymmetrical technologies are already emerging, in forms such as electromagnetic pulse weapons which take out enemy technologies, and cheap micro-satellites which destroy satellite surveillance. In Iraq and elsewhere, insurgents are pioneering the use of mobile phones as triggers, and even making moves into remote-control and robotics. How many of these measures will eventually have insurrectional uses? Already a remote-controlled graffiti machine has been created. We can expect to see the constant innovation of new asymmetrical techniques for as long as domination persists.

Raising costs

To defeat or push back states strategically, it's helpful to understand how they think. This is not easy: they think in a way which is so alien to non-renunciated life that it is hard to understand. It helps to think of the state as an instrumental machine: it functions in large part on cost-benefit rationalities. Costs are the Achilles' heel of repression. They ultimately constrain states, because they can interfere with states' abilities to pursue other activities, or the competitiveness of their capitalist tax-base. States want control, but on the cheap; and they will usually choose between tolerance and repression based on which costs more. Actually, their thinking is more complicated than this, for several reasons. Firstly, they'll sometimes bear a large immediate cost (such as the expense of the Battle of Mainzerstrasse) in the hope of future benefits (such as a smaller, more demoralised squatters' movement). Secondly, the 'cost' of the destruction of the system is for the state infinite, and justifies any cost. Thirdly, states sometimes seem to react to incalculable costs (such as moral panics) in unpredictable ways.

If done right, imposing costs allows statists (and capitalists) to be pushed back a bit at a time, cornered, disempowered, and reduced to a much less threatening position. Roughly speaking, this works as follows. If the costs are high enough, states can usually be prevented from repression. For the costs to be high enough, they need to be higher than the costs of toleration. The state may or may not choose to invest in 'speculative' repression, which aims mainly to alter the future balance of forces. It is less likely to do this, the less disposable income it has. Hence the reason insurrections are usually more effective during economic downturns. The state's reasoning will also be affected by activist responses. The less easy activists are to demoralise, the less beneficial the gamble of speculative repression will seem. Spaces crucial to insurrection can be imposed on states. States will concede a lot rather than risk collapse. Most often, these concessions feed back into recuperation. But they can also be used to carve out autonomous zones. Think of examples like autonomous student spaces in pre-neoliberal Japan, squatting in 1980s Germany and Holland, the Zapatista zone in Chiapas, de facto self-governed shanty-towns in major Southern cities, or university asylum in Greece. These are not recuperated spaces, but autonomous spaces the state was/is forced to tolerate.

We can see this statist reasoning across a number of cases. In the case of the UK animal rights movement, the state did not intervene to save various small operations such as Hillgrove Farm, but was prepared to go to very extreme lengths (from government financial bailouts to bogus trials) to protect HLS itself, viewed as central to an accumulation strategy based on biotechnology. The German squatters' movement was highly successful in the 1980s, mainly by imposing costs – a squat eviction would be met with militant protests, fierce squat defence, and the formation of new squats. This position was reversed in the early 1990s, and some cities are now squat-free. This is partly due to recuperation (most of the old 1980s squatters were legalised), partly to just-in-time policing (the tactic of attacking squats the moment they're formed, is costly, risky, often effective, but vulnerable to just one or two failures rendering it unsustainable; it is only viable because of the mass legalisations of squats and reduced numbers in the movement). Hence, the state reduced costs of repression, but also took on more costly repression – which can be made sense of in terms of rapidly rising real-estate prices in most of the affected cities.

This allows us to upgrade our sense of our own effectiveness. It must be remembered here that what seem like positive things for the state, such as jailing an activist or fencing in a summit, are actually immensely costly. States regularly spend millions on summits. The Toronto G20 summit reportedly ran to an $850 million bill. Jail costs $20,000 per prisoner per year in running costs alone. In the case of the London Mayday protests, the 'successful' repression of Mayday 2001 through mass 'kettling' cost £20 million in pre-emptive business closures alone, compared to a £500,000 total cost of Mayday 2000, deemed unsuccessful because of property damage. (The state gamble – which ultimately succeeded – was that this cost would be worth it if demoralisation and fear caused the annual protests to fizzle out). Protest often imposes costs, even when it seems to be effectively repressed. It should be added that Southern regimes often operate on a different basis, mainly because they rely on forms of repression which are less effective, but lower-cost.

This calculation on the state side can be used to modulate insurrection. Reducing the costs of autonomous activity to a point where they are small enough to be tolerated is not an option, as it increases disempowerment. There are exceptions in terms of selectivity: indirect targeting of smaller firms in campaigning against a major company, and squatting lower- rather than higher-value buildings, are two examples. Another option is to actively nibble away at a target in cumulative ways, which never cross the threshold where repression becomes cost-effective, but which add up to the collapse of the target. Usually, however, insurrection implies that ordinary action imposes extensive costs, and cutting these costs is impossible without betraying insurrection.

Raising the costs of repression, on the other hand, is viable. For this to be done, each movement needs, so to speak, capacity held in reserve. This can be achieved in two ways. Firstly, it would be helped by being less 'hyperactive', doing fewer things but doing them better, while staying ready to respond to a crisis. Secondly, it would be enabled by links between movements, such that repression of one band which was already fully-stretched produced responses from completely different bands which were not part of the same mobilisation. Hence, effective networking around issues of repression can be an effective way of preventing it. Either way, keeping in reserve a capacity to respond to repression is crucial to preventing it. Keeping up a high level of movement composition – strong connections, sustainable emotional forces – contributes to preventing repression. In Manipur, the Meira Paibis provide an example of a 'reserve' force constantly on watch for repression, something like a vastly extended Copwatch scheme, patrolling for hours each day on the lookout for state forces, ready to sound the alarm if abuse occurs. Overcommitting to the moment, at the expense of failing to keep forces in reserve to respond to new developments, impedes the ability of insurrections to handle repression. Activist bands and affinity-networks need to find ways to distribute activity sustainably through time, avoiding overcommitment and burnout.

Another way to think about insurrection is in relation to the SHAC model. This puts a particular inflection on permanent attack: there is still constant action, constant attack, drawing on expressive modalities, but it is varied in intensity and target, to increase its instrumental power. The model is often misunderstood as operating on a human level, as 'intimidation'. Primarily it operates at the level of the basic logic of capitalism, which is instrumental and inhuman: it imposes costs. Capitalists make decisions to disinvest, because the risk of suffering losses outweighs the profit which can be made. This has proven very effective in pushing HLS to the point where it can no longer function in the capitalist market. SHAC's vulnerability is that, while it imposes costs on animal abusers, it is open to retaliation by the state, on which it does not, on the whole, impose costs. It can be predicted that people will apply this kind of strategy across a range of issues, and especially, apply it to create the conditions of permanent attack: to prevent the state from repressing, to corrode its repressive capability, to carve out autonomous zones, to retaliate against state atrocities. This would in turn enhance its existing uses too, rendering the likes of SHAC less vulnerable to state repression.

One way to sustain movements in the face of repression is to turn repression itself into a source of anger, and hence of further action. This is shown in certain Southern contexts where killings by police (of activists or of ordinary people) lead almost automatically to responses: police stations attacked, mass protests called, and so on. In Iran during the 1979 Revolution, the tide was maintained because activists' funerals, held after a delay, became a site of renewed resistance, spreading into new demonstrations. In Kashmir today, when police kill, protests always follow. Even in America, police killings and deaths in custody sometimes spark unrest, such as the recent Oakland uprising. Is it possible to duplicate this kind of response in contexts where the violence used is not usually lethal? If it happened, it would probably turn a particular event (such as conviction) or the use of a particular tactic (such as 'kettling' or abuse in custody) into a trigger for protest or for other actions.

Analysing summit protests

The response to summit protests shows the strategic situation clearly. The police effectively lost in Seattle, Prague and Washington, partly because they were unable to hold space, partly because the images went against them, and partly because real disruption occurred. Police responses have followed a standardised model, and have ranged from the relocation of summits to fortified out-of-town encampments, through the use of pre-emptive arrests, “kettling”, and attacks on convergence sites, to a general increase in brutality. There are three strategic rationales to this response. Firstly, it aims to disrupt protest in general (not only militant protest), the apparent purpose being to reclaim media space by showing the police on the attack (rather than delegates besieged, or police being routed). The goal here is to hegemonise the media space. Secondly, it aims to make activists feel powerless, to disrupt devices such as the creation of temporary homespaces at convergence sites and the division of protests into zones to modulate risk, to deploy weapons designed to produce pain and disorientation, and to produce situations of frustration and sheer terror. These measures aim to break morale. Thirdly, the relocation of summits changes little in spatial terms, but reframes a forced outcome as a choice. Previously, it seemed like a defeat that summits occurred under siege, and protesters occupied the surrounding town; now, it seems like a deliberate strategy. This created dilemmas for activists. The previously highly effective 'swarming' tactic had to be abandoned. Morale-boosting symbolic victories became less likely.

This response occurred because the state did not wish to concede the space it had effectively lost with the rise of summit protests; it preferred to try to seize back this space through fascistic measures. The state thus gives up many of the deep supports of its existence, the ideology of legitimacy which disguises social war and keeps up an appearance of civil rights. This is, once more, an effect of 'just in time' policing: the state has all its forces on the frontline, and no deep support behind it; it has given up the trenches and fieldworks the maintenance of which would formerly have provided security in the event of a frontline defeat, but which restricted what the state could do on the frontline. This basic vulnerability is often missed in critiques of the effectiveness of such protests today. This said, it creates certain problems. The expected effect of such measures would be to reduce overall numbers, make it less likely that first-time protesters will attend, but also to increase militancy among protesters, who will become increasingly angry with the repression. This seems to be what has largely happened. The gap which needs to be addressed on our side is that, if such protests no longer self-recruit so easily, there is an increasing need for other kinds of bridges into everyday life, to bring new people into activism. Protests can no longer be expected to self-recruit.

Hitting the infrastructure of power

'Power is no longer concentrated in one point in the world; it is the world itself, its flows and its avenues, its people and its norms, its codes and its technologies... Anyone who defeats it locally sends a planetary shock wave through its networks' (CI)

'Every network has its weak points, the nodes that must be undone in order to interrupt circulation, to unwind the web' (CI)

'Nowadays sabotaging the social machine with any real effect involves reappropriating and reinventing the ways of interrupting its networks' (CI)

'It's within the malfunction and short-circuits of the system that we find the elements of a response whose logic would be to abolish the problems themselves' (CI)

'All the incivilities of the streets should become methodical and systematic, converging in a diffuse, effective guerrilla war that restores us to our ungovernability, our primordial unruliness' (CI)

The idea of targeting crucial nodes of power is not new to insurrectionism. According to Bonanno, because power is exercised through control over physical spaces, it can be attacked in its presence in physical space. A single act of destruction is not the same as bringing down the entire system. But multiplied enough times, it renders parts of the system unworkable. Effective insurrections often take the form of the sustained reproduction of the destruction or blockage of nodes, through time and space. Everything depends on keeping the action going, expanding it, and responding to moves to make it more difficult. There is not a qualitative difference between the small victories, tearing down all the surveillance cameras in an area or making squat eviction impossible, and the eventual destruction of capitalism and the state. The latter is an accumulation of the former, to the point where the system's functioning becomes impossible. Furthermore, if mechanisms necessary for state control or capital accumulation are taken out in this way, the state and/or capitalism would be expelled from the space in question. Social relations themselves can't be destroyed by sabotage, but they are embedded in infrastructures which can be physically targeted. The power and extent of such infrastructures affects greatly whether autonomous spaces can appear, and the costs of sustaining them make them a weak link.

There is, of course, also the question of building other worlds in liberated spaces. This process is affirmative, not destructive, and may involve quite different 'virtues', quite different forms of social relations from those involved in destroying capitalism. This needs to be done well, because the problems with the system (particularly informal hierarchies, exclusion, and patterns such as racism and sexism) are often reproduced in autonomous spaces. But this process by itself, without insurrection, could not be enough. Furthermore, since social relations recompose whenever a crisis disrupts the status quo (think of New Orleans, the Argentinazo, etc), it seems the insurrectionary part is the more difficult part. In addition, sabotage can help in the reconstruction of other worlds. Sabotage is often highly emotionally empowering. In a Black Block statement (see The Black Block Papers, p. 45-6), it is described as cracking the veneer of legitimacy, exorcising structural violence, turning limited exchange-values into open-ended use-values, changing how we see objects, increasing the 'potential uses of an entire cityscape', and breaking spells by making the impossible possible.

The Coming Insurrection makes various contributions to the strategy of sabotage. In particular, it argues for surprise attacks, which it views as central to the banlieue revolts: ambushing police patrols, attacking police stations at night and so on. In demonstrations, the equivalent tactic is taken to be bypassing the Red Zone and choosing one's own terrain. 'The important thing is not to be better armed but to take the initiative'. Another tactic suggested is opening up multiple fronts. 'Harassing the police means that by forcing them to be everywhere they can no longer be effective anywhere'.

The Coming Insurrection argues that insurrection starts with an unconditional refusal, 'a truth that we refuse to give up' – non-renunciation. This then spreads until there is victory, like the proliferation of the German squatters' movement and the French anti-fascist resistance. To this should be added Bonanno's observation (And We Will Still be Ready, 26-7) that insurrection requires replicability, not decipherability. The means by which an insurrectionary act spreads is not its comprehension by viewers, but the fact that it can be imitated and taken up by others with insurrectionary intent.

Blocking nodes

'Jam everything – this will be the first reflex of all those who rebel against the present order. In a delocalised economy [using] just-in-time production... to block circulation is to block production as well' (CI)

'the metropolis is one of the most vulnerable human arrangements that has ever existed... A brutal shutting down of borders... a sudden interruption of supply lines, organized blockades of the axes of communication – and the whole facade crumbles... The world would not be moving so fast if it didn't have to constantly outrun its own collapse' (CI)

On the longshoremen's strike: 'With ten thousand people, the largest economic power in the world can be brought to its knees' (CI)

'through the systematic occupation of institutions and obstinate blockading, the high-school students' movement of 2005 and the struggle against the CPE-law reminded us of the ability of large movements to cause trouble and carry out diffuse offensives' (CI)

Today, systemic vulnerabilities are concentrated in strategic nodes: transport and communications infrastructures (key roads, airports, high-speed rail links, ports, cellphone towers, electricity infrastructure), symbolic sites linked to capital accumulation (e.g. tourist sites), and distribution depots (e.g. petrol stations, warehouses). Targeting such sites is a growing trend among movements the world over. It is effective because just-in-time production and reduced state spending have left the infrastructure increasingly vulnerable, the system increasingly close to the wire: a small shutdown can shut down a massive network dependent on it, and have immense effects, since the system requires constant flows in the absence of stockpiles.

Just-in-time production leaves the system increasingly vulnerable to blockades. The Coming Insurrection refers to the Argentine piqueteros and the Oaxaca uprising, deemed by statists a disaster on the scale of a hurricane, and an incident in Rennes where only 300 people were needed to shut down the main access road to the town for hours. Examples could be multiplied: the struggles in Bolivia and Ecuador, the Manipur uprising, the costs imposed by blockades of timber sales, the airport and road intersection occupations in Thailand. In the successful Baliapal land grab resistance movement, checkpoints were set up on the four entrance roads to the area, and staffed around the clock. When state forces appeared, conch shells were blown and metal plates beaten to summon protesters to create human roadblocks (Routledge and Simons 488). It might be predicted that blockades will be multiplied, sustained through rolling series of blockades, used as a way to impose costs whenever the system attacks, used to defend and carve out autonomous zones. Indian social movements have pioneered a tactic known as the bandh, in which an entire local area is shut down in response to a (usually localised) abuse or grievance, complete with roadblocks, and sometimes stone-throwing. More than the workplace strike, the bandh is a strike in the full space of capital, creating autonomy in an area by shutting down 'normal life'. Something like the bandh might be used in areas where local populations are resistant to neoliberalism, to link insurrection to the wider opposition and eat away at state power. There might be a future period, for instance, in which every time American police killed a black person, the nearest city was shut down for a day. One might predict that the number of murders by police would decrease, and that autonomous spaces and feelings of empowerment would increase.

Invisibility

'turning the anonymity to which we've been relegated to our advantage, and through conspiracy, nocturnal or faceless actions, creating an invulnerable position of attack... To be socially nothing is not a humiliating position... but is on the contrary the condition for maximum freedom of action' (CI)

'And once we become visible our days will be numbered' (CI)

The theme of invisibility has a long history. James Scott's work focuses on invisible tactics of everyday resistance, some of which not only disguise the actor, but also disguise the fact that resistance has even happened. David Graeber's Fragments of an Anarchist Anthropology speculates that there are a huge number of liberated zones around the world, but most of them have stayed liberated by being invisible, and will only be stumbled across by other anarchists. Resistances usually stay invisible because this makes it harder for the state to crack down. It makes it harder to induce moral panics, or to distinguish resistances from passive effects. There are many forms of invisibility. There's the most familiar forms, such as masking-up, late-night sabotage, security culture. There's others which create an appearance of being recuperated, without actually being recuperated. There's still others which make people invisible as part of a large movement, whose members are too numerous to track down in detail.

The more militant an action is, the greater the risk that the state will turn its gaze towards it. This can often be warded off to a degree by forms of invisibility which make it difficult for the state to catch activists. Yet once visibility is established, the state may lash out in other directions, looking for a target – an innocent activist to stitch up, a community to collectively punish, a social movement infrastructure to close down as a scapegoat. When doing things which bring about partial visibility, planning should not be limited to avoiding individual detection. Preparation should also be made to impose costs on the state should it engage in repression.

Fraternising and decomposing the state

'A massive crowd would be needed to challenge the army, invading its ranks and fraternizing with the soldiers... It is not impossible to defeat an army politically' (CI)

Can an insurrection win by decomposing the state? Historically, there are cases where statists have gone over to the insurrection, as in Albania and Serbia. Yet this usually occurs when the state machine is already collapsing, and is a prelude to their hijacking of the revolt. We need to realise that statists, as long as they remain statists, are inculcated into a mentality which precludes the emotional responses necessary to identify with revolt. It is hard for compassionate people to understand the brutality of the state, and realise it is not going to decompose through statists' basic humanity. It happens sometimes with soldiers who are conscripts, or recruits from poor backgrounds (the 'poverty draft'), but it happens rarely with properly induced state agents. In the colonies, the risk of the military identifying with insurgents is managed in a simple way: they don't speak the language. In the case of the police, the same effect is achieved through 'cop culture', and often the very real blocking of communication through helmets and visors. Not to mention that, as for Crisso and Odoteo, this absense of a common language is now true for all of us when faced with the police. David Graeber argues that activists find police impossible to understand, mainly because the police's authoritarianism and the situation of conflict provide a barrier to emotional exchange. Berardi argues that the scarcity of attention available today has turned many people into ruthless executors of decisions taken without attention. Virilio argues that today's warriors are so supplemented with artificial vision that they can no longer relate on a human level. And then there's the risk that they'll simply replace human police with robots. If the state ever resorts to sending conscripts or poverty-draftees against insurrectionists, it's asking for trouble. As long as it can rely on police, robots, or tonton macoutes, it can get away with repression.

Rules of engagement

'We live under an occupation, under police occupation' (CI)

Rather than fraternisation, it is more helpful to think in terms of the transformation of the 'rules of engagement', expanding the scope for insurrectional action while restricting that of the state. In his paper on the phenomenology of giant puppets, David Graeber observes that, in conflicts between police and protesters, each side acts as if playing a game whose rules it had invented entirely by itself. In fact, the field is conflictual, and rules of engagement between opponents sometimes emerge. Aside from normally prohibiting certain kinds of deliberately lethal force, police in many Northern countries seem to recognise no limits in their rules of engagement. The reason for this is that police seek a monopoly on defining situations – they do not wish to admit the existence of an adversary.

But it does seem that the rules of engagement can be pushed in either direction. Protesters are more daring in some countries than others. Police violence is more indiscriminate from Genoa onwards than before. In practice, rules of engagement are set in two ways: in indirect effects after the event, and in impact on morale. The police have found ways to dominate certain indirect effects, notably 'bad press', through psychological operations. But this does not leave them immune to other kinds of indirect effects which impose costs on repressive actions. The ups and downs of each side's emotions are more fluid and dynamic. Insurrectional acts exist on a continuum between hope and anxiety: there is always enormous gain, in emotional self-empowerment, but this system tries to balance this with enormous risk. The level of risk varies with the countermeasures taken by the system and its ability to handle the broader context. The state tries to terrorise us because it is afraid. Though it is hard to tell when it is truly afraid, and when it simply simulates fear (to cause moral panics, for example). On the insurrectionist side, in principle rules of engagement are rejected as concessions to power, but in practice activists do hold back in all kinds of ways. The question of altering the de facto rules of engagement to our advantage – by losing our own fears, and by imposing limits on the state – may be crucial during prolonged struggles.

Sustaining Insurrection

'a blockade is only as effective as the insurgents' capacity to supply themselves and to communicate... Acquiring the skills to provide, over time, for one's own basic subsistence implies appropriating the necessary means of production' (CI)

'the state... instinctively grinds down any solidarities that escape it until nothing remains except citizenship... [The citizen] can't help envying these so-called “problem” neighbourhoods where there still persists a bit of communal life, a few links between beings, some solidarities not controlled by the state, an informal economy, an organization that is not yet detached from those who organize' (CI)

'Inhabiting a nowhere makes us vulnerable to the slightest jolt in the system' (CI)

'The destruction of the peasant's world... meant the disappearance of the means for dealing with scarcity' (CI)

The longest uprisings in the global North in recent memory have been the Greek insurrection of 2008 and the French banlieue revolt of 2005. Both of these lasted around three weeks. This has, of course, inspired activists used to four-day summit protests or even shorter upheavals, but ultimately, a month is not long enough to bring down the system. In both cases, the state largely sat out the revolt, waiting for it to fizzle out. If the state was genuinely afraid that the revolt could last forever, it could not have responded in this way. Bolivia has experienced a number of peasant shutdowns which have lasted for months. Thailand has seen protest camps which have taken over key intersections in the capital for months on end, eventually repressed by police, only to reappear a month later with similar staying-power. The Manipur uprising of 2004 was six months long at its peak. Parts of Palestine, such as the village of Bil'in, manage to continue recurring waves of protest. Argentina, Albania, Oaxaca, Ecuador, Kabylie, Kashmir... the list goes on. In other words, Northern insurrections face a pressing problem of endurance.

One possible reason movements in the South have such temporal resilience is that they are operating out of local economies which are only marginally subsumed in capitalism, and networks of everyday practices which produce a social fabric irreducible to the system: they could persist because they were really autonomous across the board, and could shut down the capitalist economy without destroying themselves. To be able to endure, an insurrection needs an autonomous economy or subsistence-system. This intersects with issues of defeating the commodity system by re-localising 'production', and expands onto broader issues in green anarchism around gift economies and ludic alternatives to work. 'Subsistence perspective' writers such as Maria Mies argue that subsistence provides a global alternative to commodity production, recognising the importance of nurturing the forces which actually produce life. The Coming Insurrection also refers to the Kabylia uprising of 2001, which effectively pushed the state out of the region. 'The movement's strength was in the diffuse complementarity of its components', irreducible to its most formal manifestation the village assemblies. The 'communes' ranged in this case from the young people fighting police to the producers of resistance symbolism and people sustaining subsistence production, without which the blockades of the commodity economy could not have been so constant and systematic.

Subsistence is also a question of producing types of bands which can be sustained over time. In a paper on precarity, Silvia Federici has argued that 'no movement can survive unless it is concerned with the reproduction of its members'. Whereas the peak of struggle today tends to be associated with events such as demonstrations, we need to be alert to questions of how to reproduce the movement through time. When communities in struggle are able to reproduce themselves – as in the indigenous movements of Bolivia and Ecuador – she argues that their anti-system struggles can become more radical. She also argues for a reexamination of the tradition of working-class mutual aid, prior to the Fordist period. These arguments echo with Hakim Bey's discussions of recreating sociality, autonomist discussions of recomposition, and primitivist discussions of rewilding. What links these fields is the creation of conditions in which insurrection can be sustained through time, which in turn, is necessary in rendering the state superfluous, and hence in destroying it. Would a successful insurrection lead beyond the current status of activism as 'bund', as entirely non-ascriptive band? This question comes down to the issue of the place of childhood in sustainability through time. Ultimately, it can be hoped that loose bands and overlapping networks can provide a context in which ascription remains redundant.

The issue of subsistence also speaks to broader issues of vulnerability. Subsistence economies operate on an orientation which favours systemic redundancy, and hence resilience, over efficiency./ Efficient systems usually produce one thing as cheaply as possible, leaving people vulnerable to shocks if what they produce is no longer in demand or if production is disrupted by social or natural crises. Subsistence economies spread their activity across a wide range of sectors, so that problems in any one sector aren't as likely to destroy the entire band. Resilience is an alternative to 'security' (the control of space to pre-empt the unexpected) in dealing with human vulnerability.

The limit in subsistence capabilities is what is holding back the temporal scope of insurrection in the North. Today's activist bands in the North do not have the degree of autonomy that some Southern movements achieve. Most often, such bands are sustained by marginal employment or state support. This is supplemented by what might be called a 'raiding' economy, taking items from the system by means such as squatting, skipping, urban foraging and autoreduction. This should not be viewed as simply an extension of the system. Indigenous groups besieged by state forces similarly rely on a mixture of raiding, marginal production and benefits. Raiding, marginality and bottom-up tribute extraction are strategies whereby systemic capture can be prevented or minimised. Yet there is a limit to how far it can sustain a movement which actually poses a threat to the system. A commune, The Coming Insurrection rightly observes, can't bank on a raiding economy forever, it needs to increase its self-organisation to meet needs. Ultimately, what prevents recomposition is dispossession: it is easy enough to live without forces work, provided one can seize back enough of what one needs.

Local knowledge

'The circulation of knowledge cancels hierarchy; it equalizes by raising up' (CI)

'Freedom isn't the act of shedding our attachments, but the practical capacity to work on them, to move around in their space, to form or dissolve them' (CI)

The recomposition of subsistence goes hand-in-hand with the recomposition of local knowledge. The Coming Insurrection recognises the need to recreate and draw on local knowledges in order to sustain insurrection. 'There's a whole set of skills and techniques just waiting to be plundered and ripped from their humanistic, street-culture, or eco-friendly trappings', not to mention 'the intuitions, the know-how, and the ingenuity found in slums'. Such techniques will have to be deployed both to 'repopulate the metropolitan desert' and to sustain insurrection beyond the early stages, in fields such as food, transport and communications. At present, our ability to attack the metropolis is compromised by our dependence on its services. Escaping this situation requires a long 'apprenticeship' in a wide range of practical skills. Communes should seek self-sufficiency, and should seek to limit their own size to prevent hierarchies emerging. In effect, what is proposed here is a recomposition of local knowledges, corresponding to a recomposition of bands.

This is excellent, as far as it goes. But it needs to go a few steps further. Band societies and subsistence economies do not begin and end with practical knowledge, though they have plenty of it. They also have very different epistemologies and cosmologies from those familiar in metropolitan societies. In practice, this always includes a spiritual element, which if examined closely, turns out to be a way of managing and reproducing emotional states. The question is not only practical but cosmological, because cosmology is necessary to sustain indefinitely the emotional states which produce insurrection. Indigenous cosmology interconnects with local knowledge, providing the frame within which it has meaning, and creating narrative structures which render local knowledge memorable and emotionally resonant. This is a situation where the truth which the band society refuses to renounce is not empty. It is, rather, the truth of a local context in its entirety. Similarly, it is not entirely the case that '[e]verywhere it's the same chilling void, reaching into even the most remote and rustic corners' (CI). There are still places where stones can speak.

“Walking through the Witches Market in La Paz - a day after the road barricades were cleared on January of this year - I realized how deep the Western view has been innoculated in my mind... I realized that my perception of reality has been modified and trained according to one model of interpretation, which standardizes the notion of the world in order to impose on us a set for socialization... In this world, life is about something else. If you cannot hear the murmur of stones there is no way you can communicate with this secret world.' - Jesus Sepulveda, 'Stones Can Speak', Green Anarchy 21

This re-localisation also changes the nature of knowledge. Local knowledge does not function like global science. Instead of recording a set of facts, it diffuses the power to create knowledge. There is much in the process of insurrection which must necessarily be a matter of situated, local knowledge and which thus, cannot be expressed in articles or books. Local knowledge has characteristics very different from those in dominant forms of knowledge in the global North. Studies of local and indigenous knowledge reveal that it usually involves a very reflexive sense of locality, situatedness, and relationality, i.e. the fact that knowledges are produced by particular people in particular places, and are relative to their process of construction and the place where they're produced. Indigenous languages tend to encourage all claims to be situated in the speaker's social position, and use words which refer to relations instead of things. (And it is quite possible that we will need to create a new language, or at least, inflect our existing language-use in ways which restore these characteristics). The Andean conception of wealth emphasises wealth in connections, not in things. An emphasis is placed on the practice of 'doing' knowledge-production, rather than the outcome. Hence, local knowledge is not a set of facts, but rather, a process of learning and sharing knowledge located in particular ways of life. It is often expressed in practices rather than communicated through books. It usually takes a holistic perspective on knowledge, rejecting the division of the world into spheres or categories, the separation of humanity from nature, and the separation of both of these from the supernatural (whether through the disenchantment of reality or the abstraction of transcendental religion). Local knowledge tends to be expressive rather than instrumental. And it tends to prefer inconclusiveness and difference to rapid decision, involving for instance long consultative processes when reaching decisions. It can be argued that local knowledge is largely a product of a subsistence economy, corresponding to a way of life which is itself situated and relational.

How can local knowledges be recreated? It is not a matter of simply importing content from other local knowledge systems – borrowing DIY skills, indigenous medicinal knowledge and so on – because this misses the importance of process in local knowledge. Nor is it about copying the rituals of other groups, or playing at being like them, which turns the immanence of local knowledge into a transcendentalism of social roles. Of course, the importation of particular knowledge-content and of techniques such as rituals can play a crucial role. Yet is it more important to recreate the generative level of local knowledge, its construction as process in an intensely situated locality. The concepts of local knowledge should not belong to the massified world, but to an intense connection to a local ecology and to those with whom one relates to this local ecology. Other aspects of local knowledge need to be recovered: an intense awareness of relationality (and corresponding rejection of 'unmarked terms' of privilege), and a replacement of instrumental orientations with a cosmology oriented to expression. Both the reifying tendencies of existing language (to focus on things instead of relations), and its pressures towards universalism and generality (towards taking one's own subjectivity as “obvious” and meanings as shared), need to be resisted.

Section 3: Dangers to insurrection

Strategies of tension

'As an attempted solution, the pressure to ensure that nothing happens, together with police surveillance of the territory, will only intensify' (CI)

The most visible danger to insurrection is the danger of repression. Its contemporary manifestation as "war on terror" follows a model of counterinsurgency shown clearly in Italy in the 1980s: creating a civil war of a type the state could use to destroy the movement under the ideological shadow of a struggle with an invisible enemy. What went wrong in Italy was not that the movement was drawn into conflict, but that sufficient costs were not levied on repression (in contrast, for example, to the pariah status which strategies of tension often bring on regimes in the South). This lack of costs has two dimensions. The first dimension is ideological: the state was able to rely on ideological gestures such as moral panics and media imagery to maintain legitimacy even while cracking down. The second dimension is the failure in the Italian case to find sufficiently effective asymmetrical means for imposing costs on repression. In other circumstances – in Chiapas, the Niger Delta, Bougainville – the strategy of tension has backfired. Research on armed opposition groups shows that repressive counterinsurgency only works with a specific kind of group, whose basis is in any case rather weak. In any other case, repression simply emboldens resistance. What's more, refraining from insurrection is no guarantee that the state will not resort to terror. And there are cases where the existence of more 'extreme' forms of opposition creates the conditions in which states are forced to tolerate less 'extreme' forms. Above all, we need to avoid aiding the state's order not to think. The point of terror is to make resistance unthinkable. Any move which aids this, aids the power of the state.

It can be argued that the global 'war on terror' is actually an indirect response to the failure of existing mechanisms of domination in the aftermath of Seattle. A strategy of tension has been unleashed which uses moral panics around terrorism and other issues to create a sense of fear which is used as a pretext to close space. This strategy plays to the psychological vulnerabilities, not only of the mass, but of activists too. A part of this dynamic is the state's attempt to 'contaminate' activism by ignoring the diversity of tactics and the division of protests into zones, adding in random attacks on less-militant protesters, bystanders, associates of protesters, people with similar ideas... In a more alert context, this could easily be met by vigorous responses. It seems, in fact, that the strategy of tension is revived to deal with each protest wave. It will fail when the fear is insufficient to curb a wave.

Moral panics

'The new economy cannot be established without a similar screening of subjects and zones singled out for transformation. The chaos that we constantly hear about will either provide the opportunity for this screening, or for our victory over this odious project' (CI)

'”Terrorist threats”, “natural disasters”, “virus warnings”, “social movements” and “urban violence” are, for society's managers, so many moments of instability where they reinforce their power, by the selection of those who please them and the elimination of those who make things difficult' (CI)

'It's useless to react to the news of the day; instead we should understand each report as a maneuver in a hostile field of strategies to be decoded, operations designed to provoke a specific reaction' (CI)

Moral panics serve as perhaps the most important state weapon today, turning what are otherwise empowering events into sources of anxiety, fear and isolation, eliminating the 'bad press' which repression would otherwise cause, and providing an enabling context for escalating repression. News coverage often functions as counterinsurgency. Police and other state agencies consciously deploy psyops to hegemonise the media field. The media complies, running police reports as fact. But there are also cases where the media targets a movement, particularly a sphere of everyday resistance, and draws the state in. Moral panics function through the dynamics of what Virilio terms 'telepresence': they focus on a single image, taken as iconic, and make the image stand for the event, at the same time turning the glare of attention on it, on condemning it, on catching the perpetrator and so on. Usually, the image is unrepresentative. Sometimes, it is created or set up deliberately by the state. Always, it is taken out of context, and used to slander entire movements. It is a major reason insurrections sometimes fail to resonate, to be replicated: they are received by potential supporters through a distorting frame.

The function of moral panics within anarchist and related movements is weaker than in the wider society, partly because people reduce exposure to, or selectively interpret, mainstream news. But there are still effects. Particularly worrying in this regard was the response to the Greek movement to the Marfin Bank incident which, while tragic, was at worst an accident and quite possibly a state set-up. The effect was to paralyse the movement, destroy the day of action and create a context where the police could storm Exarcheia with little opposition. It also led to recriminations among anarchists, and in particular, other currents turning on insurrectionists, internalising the wider demonisation inside the movement. This is not the only instance. For example, a strike wave in Korea has been defeated because support evaporated after a media scare over a video with a parodic execution of a boss. Tree-spiking went into sharp decline after an accidental death. And I suspect 911 had a similar effect on activism in America: the less composed layers were drawn into the paralysing effects of mourning, at the expense of the broader context.

Moral panic is not a tactic we can afford to recognise as legitimate. Of course we have our own ethical positions, but these can have nothing in common with systemic moralities. An incident like Marfin can only be viewed in the context of far greater slaughter and suffering resulting from the system's actions, in normal conditions of everyday life. The biggest danger here is in holding back from actions for fear of moral-panic responses. The state will not stop with one wave of panic; if the tools are allowed to work, they will be used against every form of resistance until none is left. It is always possible to deduce some risk, however unlikely, which could rule out an action. Derrick Jensen's As the World Burns contains a powerful parody of this strategy: an eco-sabotage action which might, indeed, would have killed children, if only they'd walked two miles from the nearest school and thrown themselves into the burning building. In Germany and in Britain we've seen moral panics about how police 'could have been' killed by people throwing stones, or dropping objects off buildings (despite the fact that such acts have been done a million times, without killing police). Then there's the roadblocks which 'could have' stopped ambulances, the tree-sits which 'could have' caused an agent of repression to fall while removing someone, etc. Giving in to moral panics lets the state close all autonomous space, issue by issue, band by band. Instead we need to build emotional and social barriers among our bands, which prevent moral panics from being internalised either in our psyches or our social groups. We should not be trying to distance ourselves from others deemed to be truly excluded, whereas we are the 'good' protesters. We are the excluded, and we reject the boundary between included and excluded, the state's right to select, the division into good and bad subjects.

So, what to do when a moral panic is turned on activists? The current responses fall into two categories: either to persist, to ignore the newfound visibility and carry on as before (which happens with animal rights), or to back off, holding back from action until the hostility dies down (which seems to have happened with tree-spiking). Neither is very effective. Persistence lets the state get away with persecution, which becomes more likely as activism continues as before. Backing off actually rewards the state for persecution. Yet alternatives seem limited. The third option is to meet escalation with escalation, but opponents may not have the forces in reserve to do this. A frontal confrontation on an issue is difficult, because the state can and will concentrate forces in the aftermath of a moral panic. The possibility remains, however, of sideways forms of retaliation for moral panics, disempowering or costing the state in broader ways. In particular, while it may prove difficult to get around the blockage at the site of repression, the fact of repression can be used as a trigger for actions elsewhere. This requires among other things that the affective impact of state terror be offset or transformed.

There are precedents for responses to moral panics, which involve a certain intensity of action being 'normalised', so that the response of moral panic becomes increasingly unavailable or is denied an effect. The German autonomes provide, at their peak, the main example: in the early period, moral panic was in full flow, and activists were often given long jail sentences; as the movement grew, it became impossible to suppress, and moral panic was actually reeled back, with sentences reduced and previously suppressed activities tolerated. A crucial aspect of this movement was that any act of repression, from evicting a squat to jailing activists, was met with militant protests or direct action. A similar effect might be achieved if, for instance, convictions of activists, deaths on protests, or incidents of 'kettling' led to sabotage sometime in the following weeks. If the link was clear, a cost of repression would be established.

The danger of massacre

'When things get serious, the army occupies the terrain. Whether... it engages in combat is less certain... a bloodbath... for now is no more than a threat, a bit like the threat of using nuclear weapons' (CI)

The one remaining rule of engagement, that the state normally refrain from lethal force, is rather anomalous. It does not seem to prohibit occasional statist murders which can be passed off as situationally justified or as aberrations by individual bad cops (from Carlo Giuliani to Ian Tomlinson to Alexis Grigoropoulos, to the Kent and Jackson State shootings). But it is very noticeable that the state has not so far adopted a policy of large-scale shooting into crowds. This is a historical and geographical exception. Massacre was a normal accompaniment of state repression of insurrection in the nineteenth century and perhaps up to the Second World War – think of the massacres at the Paris Commune, the 1848 insurrection, the 1918-19 German council rising, the Haymarket Martyrs, and various labour disputes in the US. And while massacre is not exactly normal in most of the global South, it certainly happens with disturbing regularity. 60 people were killed during the El Alto 'gas war', to take just one example; think too of Acteal, of Bougainville, of retaliations in West Papua, of the Argentinazo of 2001, the Peruvian Amazon protests, events in Kashmir and Palestine, not to mention that twenty years ago, some regimes (such as Indonesia) used massacre as a matter of course at every minor protest or dispute.

Why is the state holding back? It is impossible to imagine for a moment that it cares about the lives of insurrectionists or about human rights. Most probably, the state has realised that massacre has unwanted effects. Even with the current psyops dominance, a massacre generates 'bad press'. It produces further waves of protest, as people mark the funerals of the dead or turn against the police. It risks the emergence of armed opposition (the long IRA campaign, for instance, was caused by the Bloody Sunday massacre). It shows the world that social war is going on. The allergy to 'violence', the Victorian attitude to it, is notably absent in much of the global South. The current system is premised on the denial of social war, the denial that an adversary exists. It is also sometimes argued that recourse to massacre defeats the point of power, which is to rule, not to destroy: in showing that the state can crush resistance only by failing to govern, by recognising that the spirit of resistance cannot be defeated, the state effectively admits its own illegitimacy, its basis in violence. On this view, a spirit of resistance can always render a state powerless (Routledge and Simons, 493-5). Understanding this is crucial, because a successful insurrection, in which 'less-lethal' weapons had failed, would push the state to the point where massacre would be considered as an option. The state will only use the option if it believes it will work. The way to make it certain it will not work, is to make the state certain that it will not stop the insurrection this way, that it will only inflame it further. The Argentinazo and the 'gas war' were largely successful, in spite of the state turning to such measures. During the Gujjar protests in India, the police were given orders to stop killing protesters, apparently when protesters started responding in kind against police. The state's 'zero option' is not undefeatable. A situation needs to be created in which its activation would be suicide for the state.

The danger of state conspiracy

The basis of insurrection is affective and immediate. It is thus dependent on how people feel. People either get angry about something or they don't. People either feel empowered by an act or they don't. Insurrectionists can't conspire to produce effects the way the state can. The state can do things in sneaky, instrumental ways which minimise the risk it faces. For instance, it can realise cumulatively, or under the veil of recuperation, things that would spark revolt as one-off measures.

As a result of its instrumental basis, the state can also conspire to change the 'rules of engagement' almost overnight. There is a frequent problem that, just as some issue is about to be won, just as Huntington Life Sciences is about to close or summit protests are successfully holding cities, the state makes a move which can only seem unfair, altering the entire situation in which the conflict occurs. Bush famously boasted that his opponents analysed reality, but, being driven by belief, he could simply change it at will. While his opponents are busy catching up, he'll change it again. It is necessary to be prepared for such moves. At present, they have a highly decomposing effect on activist movements, particularly when repression is ratcheted up (think of the impact of Genoa and its aftermath in Italy, or the effect of the post-911 period in America). Being aware of such problems can help avert them, but it is also necessary to keep active on the current terrain, and not hold back out of fear that the state will make such moves. It is hard to come up with answers as to how to respond to this risk, but three possibilities come to mind. Firstly, that a capacity to respond overwhelmingly to escalation, held in reserve, could serve to ward it off. Secondly, that enabling the state to make other moves of partial retreat can make such options of escalation less attractive. The risk in any such preparations is that they could also enable recuperation. Thirdly, enacting similar alterations of the 'rules of engagement' on one's own side can disempower the state's own ability to change the rules. Insurrection needs to stay unpredictable, and innovate constantly.

Recuperation

'no guaranteed income... will be able to lay the foundation of a New Deal, a new pact, a new peace. The social feeling has already evaporated too much for that' (CI)

Where repression fails, there is always recuperation as an alternative. It is a quieter, less obvious way of defeating insurrections. Recuperation is an ambiguous response. It often involves real problems being addressed, sometimes even with the same responses which might occur in a liberated context, and with the recognition or tolerance of autonomous spaces. Yet it addresses such problems one by one, in such a way as to keep the system in place and hence to keep generating problems.

Recuperation follows a common pattern. The state will decide, in the event of defeat, to tolerate the costs of autonomy, because they are less than the costs of repression. But it will also try, simultaneously, to alter the future strategic balance by decomposing the basis for resistance, usually by drawing people inside on the margins. There is thus a common pattern of the state conceding something (squatters' rights in Germany, Aboriginal autonomy in Australia, a halt to road-building in Britain...), waiting for the struggle to abate, and then attacking again 20 or 30 years later.

The main threat posed by periods of recuperation is that the networks sustained by resistance fall apart, and the emotional states arising from struggle become harder to sustain. It can be difficult to retain social composition at times when recuperation is at work. There is a tendency for the number of activists to decrease, and for people to be less angry. Previous partial victories, such as the creation of autonomous spaces, then become vulnerable to a backlash after a period of time. Ways need to be found to 'lock in' such victories during periods of downturn in struggle. This could be achieved by deepening resistance in everyday life, or by radicalising the break with dominant ways of seeing.

In place of a conclusion: crisis

'What makes the crisis desirable is that in the crisis the environment ceases to be the environment. We are forced to reestablish contact, albeit a potentially fatal one, with what's there, to rediscover the rhythms of reality... [of] something to inhabit' (CI)

'Reconnecting with such gestures [of defying the state and making do with what's available], buried under years of normalized life, is the only practicable means of not sinking down with the world' (CI)

'There is a clinically dead civilization... only decision will rid us of the corpse' (CI)

Capitalism is in crisis. Maybe this is one of capitalism's periodic cyclical downturns. Or maybe it is something deeper, a crisis of the dominant system, connected to the ecological crisis, to the end of abundant energy supplies, to ecological exhaustion. The world-systems analyst Sing Chew suggests that civilisations eventually collapse from ecological exhaustion, when they run out of resources to exploit. After the collapse, there follow 'dark ages', which Chew, as a progressivist, tends to fear – but which for us, could be periods of hope. In 'dark ages', diffuse power proliferates, populations disperse from centralised spaces, local knowledge predominates over global knowledge.

Crisis creates opportunities for rebuilding. The Coming Insurrection refers to the experience of New Orleans: crisis created a space in which accumulated practical knowledges were deployed. This suggests something important: recomposition is only ever a crisis away. It has to be actively prevented by keeping the lid on the pressure-kettle, by actively decomposing. It reappears the moment a crisis makes the lid come off. They argue that movements are strongest when they take advantage for the opportunities for self-organisation created by moments of suspension of normality – from Islamic parties providing para-state services in marginal zones to left parties exploiting crises. The implication is that gaps created by crises can also be filled by activist bands, recomposing other ways of living.

Hence, we have a range of paths forward for insurrection, a range of possible futures. Certain strategies seem likely to proliferate. In particular, the warding-off of repression is likely to produce new strategies, and subsistence orientations could well re-emerge. Insurrections should be thought of in terms of counterposing a social network based on bands, swarms and affinity to a social system based on hierarchies. It is also a struggle of expressive against instrumental conceptions. The success of insurrection depends on the prevalence and intensity of diffuse power, counterposed to concentrated power. The current crisis provides a potential opportunity to expand spaces of autonomy, but certain fundamental problems would have to be addressed before Northern movements can sustain an insurrection indefinitely. Whether these problems will be addressed, remains to be seen.