The diversity of tactics: Anarchism and political power

Introduction

The Failure of Nonviolence deploys readings of empirical historical examples to show that many of the instrumentalist arguments made by pacifists and other proponents of nonviolent tactics in protest are invalid, much as Peter Gelderloos attempted to do in his earlier book How Nonviolence Protects the State (Gelderloos, 2007). The Failure of Nonviolence includes a chapter length consideration of the argument and reception of the earlier volume, in which he responds to pacifist and nonviolent critics, in particular those who disavow Gelderloos’s effort to engage with and criticise ‘nonviolence’ as a whole. Rejecting their denials that his analysis can apply to the whole of nonviolence, Gelderloos argues that although there is a broad range of groups, tendencies and positions, they nevertheless have coherence and allegiance with one another and a discernable range of political effects (307). He also concedes that the earlier book was directed at an audience of fellow anarchists and was in some places unduly harsh in the tone of its condemnation of nonviolent activists (306). In the present book, he says, he is concerned to try to tackle the question of nonviolence for the benefit of a more general audience of protesters, dissenters and campaigners, not only as recommendations for fellow anarchists (306, 307).

In this review, I try to explicate and evaluate Gelderloos’s condemnation of ‘nonviolence’ in the context of his account of political goals, strategies and tactics. The question of violence in relation to anarchist action and political power has been a prominent problem for thinkers since anarchism’s inception (Frazer and Hutchings, 2019). This book is worth serious reading by students of political thought and theory, for its sustained analysis of dilemmas of anarchist action as much as for Gelderloos’ own views about how these should be tackled.

The Failure of Nonviolence sets out three broad theses. The first is based on readings of the evidence from a wide range of campaigns, revolutionary actions, protests and movements from the late twentieth to early twenty-first century (39–119). From this evidence, Gelderloos constructs the argument that nonviolence has been ineffective in bringing about social change, whereas uses of diverse tactics, including those that cause harm and damage, and those that involve violent action, can be and have been effective. I don’t, in this essay, spend many words on these empirical studies, the detail of which will doubtless be examined, and perhaps disputed, by students of political movements and history. Second, he presents critical analysis of the sponsors, organisers and promotors of nonviolence (178–236). He reads the evidence here as showing that prominent individuals benefit materially from their professional and establishment positions and reads their normative and practical endorsements of nonviolence as a principle and as a tactic as explicitly or implicitly contributing to the oppressive and exploitative structure of governmental and corporate domination. At heart, pacifists and other proponents of nonviolence are authoritarian, so it is unsurprising that they are invested in the existing social order.

These two theses feed into a third – to which I am going to give most attention in this essay. This is Gelderloos’s normative – commendatory – theory of ‘diversity of tactics’ (10–39, 258–303). This includes a critical analysis of the conceptual opposition ‘violence’ and ‘non-violence’ and of the trio goals-strategy-tactics. In both these cases, his analysis has the advantage of shrewd insights about political language and political practice. The concept violence, it is true, is often taken vexatiously for granted, or flung about indiscriminately, or used for transparently manipulative purposes. The distinction between goals and strategy is relevant particularly in relation to the justification of philosophically and politically problematic ideals. An ideal state of affairs might be philosophically justified, for instance, as a very distant, even unattainable but nevertheless sought after, goal, while strategy should be designed only to keep the goal in view by, for instance, the articulation whenever possible of how things should be. In practical political action, this distinction is relevant to the matter of coalitional cooperation and action in concert. As Gelderloos’s analysis shows, though, the viability of coalitional cooperation turns critically on the matter of tactics. His aim in this book is to set out a normative theory of tactics, such as to make coalition between nonviolent protesters and those who do not take nonviolence to be a constraint both possible and viable.

As I go on to argue, in my judgement, this normative theory of tactics is unsuccessful in its own terms. More importantly, though, the problem is that Gelderloos offers no normative theory of politics. Indeed, in analysis that is at odds with the relative subtlety and insight of his treatment of violence–nonviolence, and goals-strategies-tactics, he treats the concepts ‘state’ and ‘politics’ very simplistically. He insists that violence is not to be treated as a thing, and quite right too (2015: 20). But the state comes perilously close to be treated as a thing by him, one furthermore that is the property and the weapon of ‘them’, and accordingly a thing to be destroyed. Politics is identified with the project of violent domination and control of state power, making ‘anarchist politics’ more or less a contradiction in terms. This rejection of politics as such robs theory and practice of the kinds of processes, relationships and institutions that can hold tactics, strategies and goals together, that can maintain a coalition across time and space.

Violence and nonviolence

Gelderloos is impatient with ethical or other philosophical or theoretical analysis of ‘violence’. The concept violence is vague – so thinkers and speakers don’t agree on its reference, having markedly different intuitions about ‘what counts’ (21). Criteria for demarcating violence from nonviolence are contested – while many will identify carrying a gun as ‘violent’ they exempt police officers from that judgement (23). He argues that this vagueness and contestation makes the category ‘violence’ irrelevant for political purposes. I concur with this judgement of vagueness and contestation, but don’t agree that they rule ‘violence’ out as a category for practical, or theoretical, political reasoning and discourse.

Let’s think about the criteria for judging an action or event or process ‘violent’. We might focus on its nature or characteristics – a wind can turn from gentle to violent, for example; a volcanic eruption can be explosive and violent, or it can be a steady and gradual flow. We are here in the realm of physical variables, in particular velocity, force, duration – that is, energy. We can also use such criteria in relation to human actions – from a (gentle) pat to a (violent) blow, from a warning uttered in a conversational tone to a shout and use of language which is loud, violent. Of course, when we distinguish between the conversational warning and the violent shout or yell, we often insert a judgement about motivation of the actor (friendliness versus aggression) or about effect on the recipient (feeling of assurance versus feeling of threat, or fear).

Such agential, motivational, effective and interpersonal criteria are independent of the event and action criteria. So there can be complex, sometimes ambiguous, cases such as the threatening whisper, which can be interpreted as violent according to the agential (motivational, effective) criterion and as nonviolent according to the action criterion. The agential, the action and the effect dimensions of the concept can be bridged, and these ambiguous cases can potentially be resolved, with the concept ‘violation’ (Bufacchi, 2007). How we analyse violation – in terms of rights, or of some other human or moral quality – is not straightforward but need not detain us here. The point is that for an action to count as violent it has both to injure and to violate. An accidental blow can injure; a deliberate or malicious blow can both injure and violate, and hence count as violence. Of course, there remain unclear cases, for instance, where intention to injure or violate is absent but the action can be judged to be reckless. Or an intentionally violent attack by a relatively weak or inept agent may fail to do any injury to a potential victim – indeed, might be hardly noticed by a stronger resilient sufferer. Violence is not always effective.

Intuitions vary with regard to whether a process or action which has the effect that a typical violent process or action has – a similar kind of injury, say – is by that fact itself violent. For some, but not for all, violence is identified by its effects. Take examples from nature – a flood or a lava flow might be quite gentle and stable and still destructive in its effects. Intuitions vary as to whether we think all such eruptions and inundations are ‘violent’ by virtue of the damage they cause in their environment or whether we should think of them as violent only if they exceed some threshold of physical energy. A more politically salient example is the idea of ‘structural violence’, according to which the injury done to individuals by the deprivations of relatively disadvantaged class locations is to be thought of as the outcome of violence just because those injuries are the same as would be done by direct forcible deprivation, or direct violent action, and they violate (Galtung, 1975). Such considerations also lead to the frequent intuitions that assassination of a vicious tyrant would not count as violence, because the all in all effects would be good, or because by their acts, or by their characters and motivations, or both, tyrants forfeit their rights or other human qualities, and cannot be violated as such.

My own view is that in the case of ‘structural violence’, it is more perspicuous to conceptualise the state of affairs as ‘the injury is as if violence had been done’, or ‘injury resulting from structural deprivation’ or ‘structural coercion’. Violation can be done, after all, other than by violence. In the case of assassination of an evil person, it is more perspicuous to speak of ‘justified violence’ than to call such acts ‘not violent’. But no matter what I think here – Gelderloos is obviously correct that there is pervasive disagreement, and vagueness, about such difficult cases. The term ‘structural violence’ has now entered the language and has a referent: the violence, the violation and resulting injury that is done by structural class position. There’s no point me denying that the term has this meaning and hence is intelligibly used in ordinary discourse.

From the point of view of normative theory, violence often connotes ‘bad’ or ‘wrong’ – it generates negative emotions like fear, effects or potential effects like injury, it proceeds from motivations like intimidation, destruction or domination, or recklessness. On the other hand, there have been notable attempts to shift this value connotation, especially in the case of revolutionary violence, or violence for justice, when it can be regarded as hygienic, aesthetically beautiful, or an epitome of virtues, or more usually all three. Notably, Gelderloos approvingly cites Fanon’s account of violence as a ‘cleansing force’ (Gelderloos, 2007: 83; 2015: 27–28; Fanon, 1965).

Anyway, despite his sympathy with the tradition of theory that contests the association of violence with wrongness and badness, Gelderloos thinks that all this conceptual analysis is quite futile and beside the point, a symptom of the failures of academic work. The point is that, because of all this indeterminacy violence is ‘useless as a strategic category’ (25). But this pronouncement is quite at odds with his own rhetorical and argumentative uses of the term. He points to examples such as oppressive policing of streets and public spaces, or the material and physical harm done by class deprivation, or the damage done to communities by corporate decision makers, as examples of the violence and injury that shows establishment condemnations of the violence of rioters to be hypocritical and self-serving. The tension is partly resolved when we observe that Gelderloos is condemning above all the strategic uses of the concept of violence by others. It is also partly resolved by his defence of the view that there is no point in activists wrangling about whether to use ‘violent’ or ‘nonviolent’ tactics and strategies.

At this point, his argument has recourse to the contradictions and hypocrisies of nonviolence. It is not only state identified and establishment assertors of law and order who come in for condemnation. The major targets of the book are counter-cultural, oppositional, protest political actors who are committed to pacifism or other variants of nonviolence. In Gelderloos’s analysis, they are guilty, as much as judges who condemn rioters while they ignore, deny or forgive police violence, of deploying the concept of violence strategically. They rely on imprecision in order to allow them to repudiate actions that they disapprove of, by calling them violent. Meanwhile, nonviolent actors tend to be invested in authority – moral authority, representation of the cause as such – which is to say in domination and the violation of other actors’ autonomy. They also rely on coercion and sometimes, indeed, on force and violence to assert nonviolence. Stewards at demonstrations and actions forcibly prevent other activists from moving and acting as they wish, and use strong-arm techniques when policing political events. (David Graeber, in his ethnography, identifies and describes ‘belligerent pacifists’; Graeber, 2009: 93.)

So Gelderloos’s argument is that nonviolence is ‘incoherent’ and also ‘pacifying’ (29). It plays into the hands of the state by proscribing effective action and permitting state violence (police brutality, use of guns, the monopolisation of public space by states and commercial corporations, the forcible prevention of its use by those who don’t comply with dominant norms of conduct, and indeed the manhandling and violent harassment of protestors and homeless people).

The ubiquity of threatened or actual state violence brings about ‘civil peace’ and compliant public conduct. Such civil peace, and dominance of politics as normal, in turn permits, indeed prescribes, the kinds of structural violence that is perpetrated by the deprivation and other ills of an unequal society.

Gelderloos also argues that to repudiate nonviolence is not the equivalent of endorsing violence (29). After all, he repudiates nonviolence in part because it is violent which itself amounts to some kind of condemnation of violence and aspiration to nonviolence. But more significantly, for him, it is because, as this nice bit of logic shows, the violence–nonviolence opposition is a flawed construct (17). As I have indicated, I go along with this, to some extent. The opposition baldly posed begs many questions – about intention and motivation, about effects, about velocity, force and duration, about agency and inter-subjectivity – which we should address using the techniques of analysis and philosophy before we can begin to articulate the criteria for calling an act, event or process violent or nonviolent.

Anarchists and political power

At this point, another argument of Gelderloos’s – about goals, strategies and tactics – is relevant. In How Nonviolence Protects the State also he denied that it makes sense to think of ‘violence’ as strategy, although both it and ‘nonviolence’ can appositely be considered as ‘tactics’. Some activists are committed to a particular set of tactics, and improperly marshal these into a strategy without appreciation of the goal; or they have a rough idea of the goal, a commitment to specific tactics, and are negligent about strategy. A problem with nonviolent activists is that they are committed to nonviolence as a strategy, which fixates them only on short-term goals. In this scheme, disputes about violence and nonviolence never get beyond ‘bickering over tactics’ (2007: 55). Goals can be ultimate or proximate. The ultimate goal is the farthest imaginable destination, so is not fixed, but will shift over time (2007: 55; 2015: 308). The strategy is the path to the goal, a ‘coordinated symphony of moves’. We should bring goals and strategies into proper relationship, and they should determine the tactics, not the other way round. Nonviolent activists severely constrain themselves when it comes to their range of tactics. And, he then goes on to argue, nonviolent so-called strategies – such as lobbying, persuasion based on taking the moral high ground or nonviolent civil disobedience – ‘all encounter insurmountable dead ends in the long term’ (2007: 67).

So far, then, in my reconstruction, Gelderloos’s normative attack on nonviolence rests on an argument from instrumental failure – nonviolence fails to achieve any goals other than to shore up the oppressive and exploitative power of state-corporate interests; and on an argument from conceptual incoherence – the violence–nonviolence distinction is untenable, a subjective and hence very variable construct which, furthermore, is strategic in the sense that it is constructed in such a way as to be maximally supportive of the conceptualiser’s own interests.

I concur with Gelderloos’s observations about the variability of the violence–nonviolence distinction, and earlier I added to the range of dimensions along which specific conceptualisations of violence and nonviolence can vary. My own view though is that first, concepts typically have ambiguous and vague frontiers because of such multi-dimensionality, and second, this is no reason to repudiate any such concepts entirely. Indeed, if we tried to dispense with vague, contested, complex or ambiguous concepts, we would be in a very restricted referential world reminiscent of the logical positivist nightmare. Second, distinctions are drawn, in context, and for specific purposes, and our drawing of any distinctions is always (or at least very often) subject to contestation. That is, conceptualisation is inherently political. Third though, Gelderloos’s problematisation of the violence–nonviolence distinction, taking his argumentative and political purposes into account, is not to be reduced wholly to its political character. His insights about political language, and about political theory, contribute to a very helpful analysis of some aspects of the concept of violence.

In particular, he shows how violence and nonviolence are implicated in one another. A straightforward view is that where one is the other cannot be. But as many who have considered the possibilities of nonviolence have admitted, or shown, the effort of repudiating violence brings one into relation with one’s own and others’ violence in ways that those who are not effortfully striving to maintain a bright line don’t confront. Now, ethically, many pacifists – Gandhi for instance – insist that this encounter with violence and deliberate turning away from it bespeaks a virtue – courage, truthfulness – that is not embodied by those who deal in degrees of violence and nonviolence with more indifference (Frazer and Hutchings, 2015; Gandhi, 1986: 298, 213). Gelderloos’s ethical view is that the preoccupation with nonviolence – as, indeed, a preoccupation with violence – leads one into authoritarian ways and relations with others.

I am going to come back to this ethical question. For now, though, I want to turn to the question why, given this insightful, and ethically and politically interesting, deconstruction of the clear violence–nonviolence binary, Gelderloos treats the significant concepts ‘state’ and ‘politics’ so simplistically. No sophisticated view of their constructed nature or ambiguity here. ‘The state’ is a ‘millennia-long movement towards centralisation’; states come to power and domination by killing (261). States make societies unsafe; the state does not exist to protect us (14, 299). The state is treated as an entity, or at least like one, albeit also as decomposable to a specific set of individual persons in jobs and positions that give them a direct interest in violence to, and exploitation and oppression of, others, whether on their own behalf or on behalf of their employers and bosses.

Politicians, meanwhile, engaged as they are in office, institutions and the competition for the power to govern using state violence, are unambiguously enemies. ‘The worst kinds of human beings tend to be attracted to the power that inheres in the role of politician or cop, along with a few people with very naive ideas about how to change the world’ (267). Politics is identified with the project of violent domination and control of state power, while revolutionary action is for the defeat and destruction of state power, for the achievement of individual autonomy preserving and human need satisfying communities and enterprises. This understanding of politics as coextensive with state-centred violence is common in anarchist understanding. It begs the question, obviously, about anarchists’ own ‘politics’. In Gelderloos’ account, anarchist action in this context of state violence mainly consists of fighting back – not because anarchists want to fight, necessarily, or valorise violence, but because there is war, there is a context of state violence, so defensive and aggressive violence is necessary (2007: 63, 69; 2015: 299). Indeed, the state will particularly target and pick on those who dissent from or try to resist state and corporate power, in particular anarchist defenders of individual autonomy.

Both these conceptions, of an unambiguously centralising authoritarian and violent state and of politics as the competition for the power to wield state authority and violence, are obviously problematic. Both concepts, for most thinkers, are complex and multi-dimensional. The boundary of the state – whether that is thought of as the state-society boundary or whether the boundary is between state and (reactionary, compliant) society versus revolutionary resisters – inevitably must be vague, ambiguous and shifting. Gelderloos’s discussion, actually, shifts on the matter of how clear the state boundary is. Much of the time he insists, or takes for granted, that there is a very clear distinction between on the one hand the state and its agents who wage war on and use violence against those who will not comply with its norms, laws and authority, and on the other hand the societies and potential communities of relatively poor and disadvantaged people who are dispossessed by the corporations the state defends, and oppressed by state violence (2007: 68).

However, society, including oppositional society, is pervaded by groups and interests who are really on the side of the state – institutions such as NGOs, political parties, trades unions, the media (2015: 267). These have to be rejected, because their centralisation and organisation are autonomy destroying, in the same way that state laws and regulation are. There can be no reasonable discussion with institutions – you can’t converse with an institution and when you speak with a member you are talking with someone who has surrendered their own discretion and is ‘performing a mechanical role’ (267–268). The difficulty, evidently – and one that in my view Gelderloos doesn’t satisfactorily address or resolve – is how this conflict within society, between the revolutionaries and the reactionaries, is to be managed by anarchist individuals. Even if an anarchist believes that anyone who is a member of or a representative of an institution is performing a mechanical role in the interests of the state, it can hardly be the case that treatment of them should be exactly the same as treatment of the police or military. Given that so many institutions are themselves the object of state power, and identified ambivalently by state and governmental officers as both legitimate and illegitimate, a clear ‘them and us’, friend and enemy distinction cannot be stably maintained, any more than can a very clear state-society frontier.

There is a pluralistic and diverse range of oppositions, protests, dissenters from and challengers to state authority. As Gelderloos puts it, ‘Not everyone who marches together in a protest is on the same side of a given social conflict’ (260). The answer is to concentrate on individuals – on those who participate in actions, those who recognise one another as fellow participants, those friends and comrades who influence one another daily (268). There can be no central organisational pivot or authority. Gelderloos articulates a vision of possible anarchist, and effective, revolutionary action. The aim is to take over spaces, to defend them, to defeat the police, to shut down the economy, to attack structures of domination, to organise health care (264). Activities might be violent or nonviolent – individuals make their own decisions about how to protest, and they think about how those decisions will affect others (274). Care must be taken with individual small businesses and neighbourhoods, but financial districts are not neighbourhoods.

Gelderloos insists that combative and peaceful tactics can work together. Reflecting on his experience in Barcelona in 2012 he observes that ‘the crowd contained a wide range of niches and possibilities for participation’. While the front line rioters burned a Starbucks, a bank and a shopping mall, and fought a pitched battle with the police that lasted hours … there were also ‘old folks and families with children, and … people cheering the rioters and booing the cops, people helping take away the injured … people arguing with pacifists’ (287). Riot, Gelderloos concludes, ‘provides a model for a stronger form of action that has a place for everybody, as long as they accept the legitimacy of other kinds of participation and reject the attempts of the police to dictate how we take over the streets’ (288).

This is a vivid sketch of anarchist action, based on mutual respect and forbearance between revolutionary participants. Let us grant that were such actions to occur in many, or a critical number of, locations simultaneously they could constitute a genuine disruption and destabilisation of the capitalist-corporate-state system that is critical to our current patterns of inequality, deprivation and injustice. Still we might be sceptical on several points.

First, this mosaic of tactics, as Gelderloos himself argues, as I cited earlier, doesn’t add up, by itself, to a strategy. It is not clear at all how the diverse goals and commitments to strategies of different actors, actually will interact with one another in this kind of revolutionary situation. Second, it is not clear how incipient centralising organisations are actually to be excluded from any such revolutionary action. The state is represented by police, by security guards defending corporate premises and by other employees of the corporations which are at the centre of capitalist inequality. Here, that is, there is a clear bright line, which separates ‘we’ revolutionaries from ‘those’ defenders of the exploitative state. Any uncertainty or vagueness of this line is to be resolved by locating organisations, representatives, politicians and others firmly on the side of the police and the capitalist corporations.

The first difficulty is how such others are to be identified: if anyone who is a member of, or representative of, an organisation or institution or who speaks on behalf of others is irredeemably tainted by their identification with state authority, the crowd inevitably must be constructed suspiciously and warily by anarchist activists as permeated with the enemy. Even if anarchists can tell who’s who, the proscription on any degree of treating with such parties, means that anarchist action must be closely policed. At best, experienced revolutionary activists can use their enemy detecting intuitions and can withhold solidarity from those who are identified as enemies of the revolution. At worst, the suspicion and antagonism between protest groups will squeeze out the solidarity that might bring gains for disadvantaged groups. All of this matters because these matters of strategy and tactics are as important as Gelderloos tells us that they are.

The simple view of state versus revolutionary actors makes anyone who dissents from the use of injurious, destructive or other violent tactics into an enemy. That is pretty breathtaking, particularly when we compare it to Gelderloos’s relatively complex and subtle analysis of ‘violence’. Gelderloos calls for all to respect and forbear from interference with those who choose to injure, destroy or use violence. He intends that this forbearance, and respect for each others’ choice of style of action and use of tactics, is properly linked to the strategy for enabling and perhaps reaching an anarchist goal of empowered local communities and individuals freed from the tutelage and violence of corporations and states.

But there surely are good reasons to dissent from the view that the tactic of riot does really clear the way for anarchist community. We might dissent from the principle that attacks on individuals who have jobs that put them on the side of ‘the state’ are fair because they are fair game. Similarly, Gelderloos’s accusation to nonviolent activists is that they are authoritarian in their imposition of their own norms and standards on all. His answer is to exclude nonviolent activists from anarchist solidarity and by implication from the realm of care and concern.

Gelderloos certainly would not be sympathetic to the view that this ethical problem rests on a political mistake – on a mistake, that is, about the concept of politics, what it encompasses, what it can mean and connote and the ethical attitude to human beings that is entailed by it. Of course, the concept politics is contested. To be sure, you don’t have to be an anarchist to go along with the view that anyone who chooses politics is by that fact considered to be ethically suspect, mendacious, manipulative, unprincipled and too ready to be the agent – at suitable arm’s length – of state violence. The theory that politics is war, that war is the basic structure of unequal and exploitative societies, and that therefore we are in violent contexts in which uses of violence are nonoptional, is also one that is shared beyond anarchism. For many thinkers and critics, politics is indeed invariably a matter of the securing and maintenance of party and individual advantage in the extended competition for the power to govern. Political actors’ ethical values are compromised by their commitment to strategic reason and action. At worst, strategy can be for evil ends or for evil ends that are cunningly disguised as good ones. All these are stereotypes of politics and politicians that pervade many political cultures. Just as I accept the ordinary meaning of the term ‘structural violence’ while disputing its conceptual coherence and cogency, I accept that these, in our world, are established meanings of politics.

But they do not do justice to the concept or the phenomenon, any more than Gelderloos’s simple identification of the state does justice to the formations of power and authority in our worlds. There are other, productive, ideas of politics, including those that centre on the idea, the ideal, of coalition. Gelderloos, rather grudgingly, accepts the necessity of coalition but refuses to follow through the political logic of the idea. Anarchists’ ultimate goal is ‘no polity’. Strategically and tactically this leaves them in the awkward position of denying the political nature of their own processes. But the kinds of norms and distributions of responsibility that characterise anarchist groups’ deliberations and planning – for instance, those described and analysed by Graeber – are clearly political in several senses. Arguments and deliberations about what is to be done, and how, are infused with reflexive arguments about how it should be decided what is to be done and how. They are infused with second guessing about other participants’ real motivations, agendas and identities. Decisions are made, but they are never final although some of them are fateful and generate unintended and unforeseen consequences (Graeber 63–89, 124–128) and so on. Politics, political institutions and organisation, political relationships and processes, put us uneasily between strategic rationality and uncertainty, between final decision and endless revision, between instrumentality and the pulls of principle and commitment.

These considerations apply, actually, even to those who construct ‘politics’ as a much more solidaristic, concerted, decisive form of collective action than this. We might think, for instance, of Arendt’s conception of public action, balanced between competition and concert, which is the only way to make a world and to construct institutions that will endure into the future. Or, we might think of pictures of legitimate sovereign policy making under conditions of autonomous agreement of each with all to be bound by laws one has had a hand in making. We might think of the Machiavellian rule of conciliation: The need to find the common ground that allows us to go on together even while we compete for the power to govern, or oppose each other’s views of what the polity should be. Politics is also the process of action in concert, in which we live with the product of diverse actions oriented to our shared life. In each of these cases, political decision is balanced between finality and openness to revision, political institutions are both constitutional and outside the constitution, political relationships are both fragile because contingent, and robust because based on shared commitments and hopes.

Politics requires tolerance of ambiguity. It also critically relies on our capacity for coalition – for working with others who share one’s purposes only to some degree. That must apply to anarchist groups, networks and aspirations as much as to anyone else. As an anarchist theorist, Gelderloos is concerned above all to safeguard individuals’ dissent and exit or noncompliance. He draws us a picture of anarchist activists in fleeting coalition with other activists without any institution of such coalition. The problem with this is that something is holding the anarchists – the affinity groups, the comrades, those who recognise each other from the activist training and the working groups – together, and that something is, at least in part, a political tie. Gelderloos’s account of anarchist action doesn’t acknowledge that the relationships between anarchists are political. Insofar as they are politically tied, anarchists participate in commitment to the future. The logic of political relationships means that actors have an interest in each other’s conduct that is at odds with Gelderloos’s picture of forbearance. His account of individual decision making and action can’t get to grips with the way that the conduct of some affects all, and the way that this is the dilemma that political engagement is precisely fashioned to deal with. His simple identification of politics with state authority and coercion disavows all ambiguity, as does his separation of actors into friend and enemy. He exploits the complexity and ambiguity of ‘violence’ argumentatively to disavow nonviolence as a principle. But these treatments of the concepts state and politics amount to an avowal of what is surely a serious kind of violence: the ejection of individuals who are not full comrades from the realm of care and concern.