Policing Atmospheres: Crowds,Protest and ‘Atmotechnics’

Affective Atmosphere and Atmotechnics

Most people will be familiar with the experience of a particularly intense affective atmosphere: the anxious hush of an exam hall, the throbbing heat of a nightclub, the strained tension of an Accident and Emergency waiting-room. Unfortunately, despite this familiarity, until quite recently affective atmospheres have not been particularly well examined. In fact, as Gernot Böhme writes, they present a fundamental difficulty:

We are not sure whether we should attribute [atmospheres] to the objects or environments from which they proceed or to the subjects who experience them. We are … unsure where they are. They seem to fill the space with a certain tone of feeling like a haze. (Böhme, 1993: 114)

In recent years, there has been a growing interest in affective atmospheres. Studies of affect have proved particularly fruitful (Anderson and Ash, 2015; Anderson, 2014; Blackman, 2014; Berlant, 2011; Ahmed, 2010; Blackman and Venn, 2010; Brennan, 2004). From affect studies we might say that atmosphere is a plane upon which certain affects circulate and commingle. It is the tone of a space, its mood, ambience, sense or feeling. For those within this atmosphere, this circulation of affects may generate particular bodily responses (endocrine, neurological or dermal changes) which might, in turn, evince particular feelings or emotions (Brennan, 2004; Gibbs, 2011). Or they might remain simmering, affecting one's response to various events, without ever reaching consciousness. In this sense, affects cannot be reduced to emotions (Halsey and Young, 2006). They circulate between bodies, sticking to them in different ways (Seyfert, 2012: 29–30; Ahmed, 2010: 36). Ben Anderson argues that:

the term ‘affect’ has been used to describe a heterogeneous range of phenomena that are taken to be part of life: background moods such as depression, moments of intense and focused involvement such as euphoria, immediate visceral responses of shame or hate, shared atmospheres of hope or panic, eruptions of passion, lifelong dedications of life, fleeting feelings of boredom, societal moods such as anxiety or fear … amongst much else. (2014: 5)

Affects can also be embodied in entities other than human bodies and circulate between these ‘bodies’. As Eve Sedgwick writes, they ‘can be, and are, attached to things, people, ideas, sensations, relations, activities, ambitions, institutions, and any number of other things’ (2003: 19). They ‘are constantly infusing embodied practices, resonating with discourses, coalescing around images, becoming part of institutions’ (Anderson, 2014: 6). So, for instance, Anderson shows that we cannot begin to think about ‘hope’ shorn of its site of circulation and mediation: the rescue of the Chilean miners in 2012, the ‘hope’ that Obama seemed to bring upon his inauguration, and the sentiment of ‘hope’ that is measured by the consumer prices index (2014: 1–5). We must attend to the ways in which spaces, environments, and social contexts might be ‘tinged’ or ‘tinted’ with a particular mood or ambience.

I want to suggest that it is useful to distinguish between affective atmospheres and the techniques aimed at creating them – what I call ‘atmotechnics’. Atmotechnics are the techniques that aim to create, manage or change affective atmospheres. In most cases these techniques are deployed to try to manage people's comportment and behaviour. We often experience strong atmospheric dynamics in everyday spaces – streetscapes, supermarkets or restaurants. Chris Hudson's atmospheric analysis of the ION Orchid in Singapore instantiates this perfectly, drawing out a variety of different techniques (from the spatial dynamics to the managed ‘aromatopias’) used to change people's comportment (Hudson, 2015). She concludes ultimately: ‘Place atmospheres are created, evoked and manipulated to facilitate the social practices expected of consumers [in the context of shopping malls] and to modify behavior’ (2015: 304). This is different from Sloterdijk's deployment of the term ‘atmotechnics’ (2002, 2016). For him the term signifies the discovery of the ‘environment’ as something that can be managed through air-conditioning. In this article, we will focus specifically on the manner in which atmotechnic interventions seek to shape the capacity to act by changing the affective atmospheric dynamics.

I use the term atmotechnics for two reasons. Firstly, there are significant epistemic and ontological difficulties with affective atmospheres. Atmospheres have an ineffability; they tend to remain in the background and cannot be reduced to conscious perception. For many, they are ‘nonconscious, noncognitive, nonlinguistic, noncoherent, nonrational and unpredetermined’ (Gould, 2008: 23), and so they are resistant to naming. While Anderson and Ash (2015) have proposed various ‘atmospheric methods’, by switching attention to the atmotechnic interventions rather than the swirling atmospheres we can bracket the questions of atmospheric ontology. As we will see, this bracketing is not complete as it is impossible to discuss atmotechnics without a sense of how the technical interventions shape atmospheres. However, it does allow us to engage with atmospheric techniques without having to agree, for instance, on whether there is only one single internally differentiated atmosphere (Philippopoulos-Mihalopoulos, 2015) or whether there are plural overlapping atmospheres (Anderson and Ash, 2015). It is possible to investigate the atmospheric techniques without displacing the importance of investigation into atmospheric ontologies.

A second reason to distinguish between atmospheric techniques and atmospheres is that it is important to be able to identify the intentional attempts to manage or change the atmosphere, and to see how the atmospheres that are actually produced are always contingent. While engineers, designers, architects and those employed to manage the space might configure it in particular ways, the atmospheres that emerge will always remain just beyond control. There are always too many possibilities, too many ways in which the entrainment of bodies in space might be distorted or unsettled. For instance, we might think about the atmotechnic management that can be seen in airports and other highly securitised spaces of transit. Departure from London's Heathrow Terminal 2 requires that you pass through four key atmospheric zones: the check-in zone, the security zone, the duty-free zone and the departure gates. The security zone is differentiated from the others – you enter through a narrow passage with darkened walls. There is less natural brightness so the 10 flood-lights atop each of the 12 structural pillars provide a more intense, bright, sterile white light. The space itself is rectangular and without ornament. Despite the crowded queuing spaces, the soundscape tends to be muted. There is no hubbub; instead, the primary sounds are the electronic beeps of the large security machines. Security agents embody a stern directness towards the passengers. In the space people tend to become more biddable, focused on getting through it as quickly as possible. Importantly, the atmotechnics change in the ‘duty-free’ zone. One enters on a mezzanine and the cavernous space of consumption opens out before you. It is bright, and the far wall is three storeys of glass. The sterile sense of focus embodied by the security zone is replaced with an ambient encouragement to tarry and consume. This is an atmotechnics – a careful planning of space, demeanor, environment, facial cues and a myriad of other factors to generate a particular atmospheric dynamic. However, we could imagine a myriad of ways that this might be disturbed – from the more mundane event of crowds being forced to queue for hours because of understaffing, to a sudden mass passenger protest or the orders to evacuate the zone due to a fire or security threat. In these scenarios, the normal atmosphere of the zone is disrupted. The atmosphere becomes imbued with the irritation, anger, antagonism, fear or flight of the travelers.

Ultimately, there is a cyclical relation between atmosphere and behaviour. Atmospheres shape people's comportment. They function as an invisible pressure on behaviour. However, at the same time people's comportment, demeanor and behaviour play an important role in shaping the atmosphere. This is particularly the case in protest, when crowds intensify this loop between behaviour and atmosphere. Thus, the spatial dynamics of a cathedral, supermarket, airport or stadium lend themselves to particular atmospheres (cf. Bitner, 1992; Zumthor, 2006; Hudson, 2015), but can always be performed differently, particularly if a crowd gathers. A quiet library might be disrupted by noisy protestors; the flow of traffic on a major road might be stopped by people sitting down; a noisy street might be disturbed by a large, somber, silent march. Space does not determine the atmosphere, but certain spaces lend themselves to particular atmospheres. This can be intensified. So, for instance: certain types of artificial light are particularly atmospheric – which is to say they are conducive to particular types of intense atmospheres; light sources such as burning torches, bonfires or burning cars can be contrasted with the sterile atmosphere engendered by police flood-lighting; particular smells can evoke different moods – the burning rubber of that car, the hint of tear gas, the sweaty fug of damp bodies; or sounds – the rhythmic chants and beats in generating a sense of commonality within the crowd (Applegate, 1969; Le Bon, 1903; Sloterdijk, 2016). All of this is to expand on my initial observation that atmosphere is the tone of a space – its mood, ambience, sense or feeling.

But in introducing the question of atmotechnics we can also begin to see its fuzzy edges, where there is not a clear line between intentional atmospheric engineering and actions, like Pussy Riot's ‘punk prayer’ performance at the Cathedral of Christ the Savior in 2012, which are undertaken for different reasons but which have very clear atmospheric effects. The tone of a space may be conducted by the design of the space or by actors who aim to shift the dynamics with particularly honed interventions. But it can also be created or shifted by activity that is not informed by specific analysis of atmospheric dynamics. Thus, while I claim ‘atmotechnics’ is a useful tool of analysis, I do not for a second suggest that it allows us to circumvent the question of atmospheres. With this caveat in mind, let me begin to detail public order atmotechnics. These are techniques that are designed to change an atmosphere, thereby altering a crowd's demeanor or behaviour in order to regain control over bodies or space.

The Atmotechnics of Escalated Force

We can exactly pinpoint the first full deployment of officers trained through the ‘escalated force’ model of the 1983 public order manual. On 18 June 1984 around 5000 picketing miners travelled to the Orgreave coking plant to prevent strike-breaking lorries from bringing coal to the British Steel mill in Scunthorpe. What the miners could not have known in advance was that, for the first time on British soil, they were facing a ‘third force’ – neither the simple bobbies who had failed to restrain the miners at the Saltley Gates in 1972, nor the army that had been on the streets of Northern Ireland since 1969. ACPO's 1981 decision to borrow large swathes of Hong Kong's public order manual led to an entirely secretive training programme, in which the tactics of colonial policing were taught to approximately 20,000 officers. This ‘third force’ was the embodiment of the ‘escalated force’ model. The extensive violence of the police that day has been the subject of much political, legal and academic consideration. However, there has been no engagement with the atmospheric techniques in the police's ‘show of force’.

‘Escalated force’ is a model that deploys police as a threat of force against protestors (Schweingruber, 2000). If protestors fail to disperse or cede their ground, the police progress from a show-of-force to the actual use-of-force. One of the key US texts on ‘escalated force’ was Rex Applegate's Riot Control. There he explained that:

The Show-Of-Force can be based on either the surprise appearance of a large unit of riot-equipped police officers in a formation, or by using the saturation technique, massing large numbers of personnel at a given point in full view of the mob. To crowds, the psychological impact of witnessing the arrival of large numbers of police emerging from patrol cars, special riot vans, police helicopters, etc. can be very great. (1969: 35)

The Show-Of-Force … may so impress the crowd … that no further physical deterrent action is necessary…. The surprise advance [from an obscured position] on the mob is made in the direction that dispersal is desired. A column formation is generally used in the approach phase, followed by a deployment out of mob contact into a wedge or line so the police and their armament can be fully viewed by the mob. (1969: 36)

The Commissioner of the Royal Hong Kong Police, Roy Henry, would later explain this tactic in slightly less analytic terms: public order policing was ‘the projection of police units in an efficient, effective and formidable manner which creates an atmosphere in the riotous mobs of apprehension and awe which could be close to fear’. If this atmospheric projection is successful, ‘the crowd will scatter: “They run like the dickens!”’ (Northam, 1988: 136).

We might group a large number of the tactics described in the Public Order Tactical Operations Manual (ACPO, 1983) as elements of the overall show-of-force. From the defensive and offensive formations of police bodies to the descriptions of dog handling or baton charges, each contains both an atmospheric element (show-of-force) and a physically forceful element (use-of-force). However, there are a number of tactics that are entirely atmospheric. For instance, section 17 of the 1983 manual describes a tactic where police beat their shields with truncheons – a technique deployed at Orgreave. The description of the tactic reads:

despite training, confidence and levels of suitable equipment, police officers deployed against hostile crowds during public disorder are likely to experience emotions ranging from anxiety and fear to outright anger. The use of chanting, shouting or the rhythmic beating on protective shields can act as a morale booster prior to deployment and also serve to release stress in police officers. (Northam, 1988: 89)

The manual explains that one of the advantages of this tactic was the loss of ‘police … anxieties about impending deployment … in the sense of group confidence engendered’ (Northam, 1988: 89). Chants, songs and other shared audible cues help establish a sense of togetherness within a crowd (Applegate, 1969). The rhythmic noise was supposed to steel the police against the protestors and valorise their sense of themselves (cf. Sloterdijk, 2016: 384–98). The manual continues: the ‘noise levels draw attention to a police show of strength’ (Northam, 1988: 89). So the drumming is also understood to radiate outwards, affecting the protestors by drawing attention to the strength of the police. The drumming is meant to encourage the crowd's withdrawal without the need for the police to apply (physical) force. It is supposed to make the crowd feel afraid, to disperse them because of the sheer menace. Failing this, the awe, fright and trepidation created might increase the efficiency of the physical force that the police subsequently deploy.

In the literature on protest and public order, this type of tactic is discussed in terms of perception. The drumming is significant insofar as it encourages protestors to imagine the police as threatening irrational violence, particularly where a protestor perceives her actions and the actions of those around her as reasonable and measured. The actions of the police are ‘read’ in a particular way by the protestors, onlookers and the media. The task for this type of analysis is to begin to understand this socially constructed symbolic language of events and plug that back into our extant knowledge of psychological processes. However, I want to suggest that an affective analysis is also important. We might think of the dread-excitement that a protestor would feel in the pit of her stomach, that strange tension that unsettles her, passing through each of those around her as they wait for ‘something to happen’. The drumming is a material communication; the beat of many officers’ batons on shields is a powerful vibration of material. The vibration reaches the ears of our protestor, but it also reaches her stomach, chest and skin. Her body changes as it accepts the message. Endocrine processes change, glands release hormones, her body produces chemicals that change the bodily processes themselves. What's more, these processes are not contained by her body. From her skin and on her breath chemicals are released that are received by the bodies of those around her (Brennan, 2004), and vice versa. In this way the response of one person is only the beginning, as different endocrine responses are communicated between bodies. The affects are contagious (Tarde, 1903). At the same time, there are mimetic affects, looks of fear or anger that are picked up (Redekop and Paré, 2010). And this is not just about the response of the protestors, because of course police bodies are undergoing similar changes. In short, the drumming is not a simple symbolic communication from the police to the protestors. Instead it contributes to the shifting and changing affective atmosphere, in part being created by the atmotechnic tactic.

We can contrast it with another tactic discussed in the 1983 manual: ‘Recorded sound or music, preferably of a soothing nature or consisting of popular songs [may be played] to induce members of a crowd to begin singing’ (Northam, 1988: 90). Like the drumming, this is an interesting tactic because it is deployed primarily for its affective power. However, unlike the drumming, this is designed to de-escalate the protests. The manual points out that this de-escalation of the atmosphere might affect the crowd which could sing along, but it also can have a calming effect on the police present on the scene (Northam, 1988: 90-1). Thus, with this technique the police commander intervenes in the affective atmosphere in order to change the type of comportment, behaviour and conduct of the crowd and also of his officers. The manual continues that the tactic ‘is of little value once actual fighting, arson or looting has begun’, and the author mentions it was treated with derision by users of the manual (Northam, 1988: 91). Escalated force has clear affective and atmotechnic aims. The calm that this music seeks to evince only underscores the aim to excite fear and awe in the rest of the show-of-force atmotechnic tactics.

The Tone of Strategic Facilitation

There are crucial atmotechnic novelties in the public order policing models adopted in the UK in the 1990s and 2000s (‘kettling’ for instance). However, let us focus on the current UK approach of ‘strategic facilitation’. The police began to reorient the UK public order training programme following the failures of policing the G20 protests in London in 2009. Her Majesty's Inspectorate of Police (HMIC) produced the key text Adapting to Protest, which drove these reforms. The cornerstone of HMIC's proposed reforms was the adoption of the Elaborated Social Identity Model (hereinafter ESIM) as the basic psychological analysis of crowds. While this model works entirely from the question of social identity, the reforms based upon it have crucial affective and atmospheric dimensions. The ESIM originates in the work of Steve Reicher in the 1980s (1982, 1984, 1987) and then extended and elaborated by Clifford Stott and John Drury (Drury and Reicher, 2000; Stott and Drury, 2000; Reicher, 1996a, 2001). The basic premise of the ESIM is that crowds generate new collective social identities, which have an important impact upon how they understand the actions of other parts of the crowd and the police.

In a number of crowd events demonstrators have come to an event as liberal subjects who want to express their views and who see the police/state as a neutral guarantor of their rights (including the right to demonstrate). The police, however, often due to the presence of some radical groupings, have viewed and treated the crowd as oppositional: they have been prevented from going where they wanted or voicing their demands to their intended audience. Such common repressive experience has led crowd members to become more united, to see themselves as more oppositional, to give greater heed to the voices of radical groups in their midst and feel more empowered to challenge what they see as illegitimate police action. This, in turn, has confirmed the original police understandings and so the process of change has escalated. In short, people positioned as the opposition have become oppositional and have reconstructed the nature of the world from that vantage point. For ESIM, self-categories are not about perception, they are both reactively and proactively about the organisation of social practices. (Reicher et al., 2010: 60)

For public order policing, the key observation of the ESIM theorists is that any time an out-group reacts aggressively, or is perceived by the in-group to act aggressively given their imperfect knowledge, the response is often to strengthen the in-group's common identity. In turn this may increase the chance that the in-group will respond with aggression to the actions of the out-group. So, in a public order setting, if the police are seen to put on riot gear, to unfairly beat or gas protestors, then this may strengthen the protestors' resolve by encouraging a common feeling of the protestors as a coherent unit.

The ESIM emphasises a process of symbolic identification of those within a crowded protest. If this symbolic logic was the only causal factor in crowd behaviour, then my affective and atmospheric analysis would be precluded in this instance. Where the ESIM relies upon a representational logic (Reicher, 2011), affect is a material communication (Massumi, 2002, 2015) with a complicated relation to representational logics (Leys, 2011; Mazzarella, 2009; Anderson, 2014). However, ESIM studies often uncover descriptions of affect once their subjects begin to describe the experience of crowded protest. We might look, for instance, at Reicher's superb study of the 1988 ‘Battle of Westminster’ student protests. There he quotes students and police alike, talking about hostile atmospheres (Reicher, 1996b: 124), instinctive communication of bodies (p. 122), types of overbearing or affect-generating presence, as well as affects of fear, anger and hostility (p. 123) and empathy (p. 126). In Reicher's analysis, these affective encounters and atmospheres get sidelined. They are secondary to the common social identities of those involved. However, I suggest that it is important to pay attention to affective descriptions in order to understand the manner in which social identity and affect coalesce. Thus, while the ESIM emphasises representational logics to explain its efficacy, it cannot help but notice affective and atmospheric dynamics.

The tactics that flow from the adoption of the ESIM are ‘centered upon the facilitation of peaceful behavior’, increasing ‘police capability for dialogue and communication with crowd members’ and ‘a graded, differentiated and information led approach to police use of force’ (HMIC, 2009: 89). The attempt by the police to surprise protestors with an overwhelming projection of might, as seen in the ‘escalated force’ model, is pushed to the background. Instead police are encouraged to deploy the low-key presence of liaison officers, and a ‘no surprises’ approach to protestors. This includes revealing key strategic decisions to trusted groups in advance of the event. The ‘Protest Liaison Team’ are central to this. They are modeled on the Swedish ‘dialogue officer’. Their function is to build the trusting relationship with protest groups over an extended period. These relationships are supposed to help the police establish a calm setting for the protest and encourage ‘self-policing’. Thus, the police can hold out the carrot of low-intensity policing but there is also a stick. Where a protest liaison team is unable to establish a trusting relation with protesters, this can be taken as evidence of the need for a more forceful model of policing.

At the heart of the protest liaison team function is the generation of an affective feedback loop. The College of Policing's manual for protest liaison team trainers (2016b) is instructive. It makes clear that their role is not simply to encourage verbal communication, but also an affective communication of bodies. Trainers are asked to explore the different meanings of ‘communications’ with their students (College of Policing, 2016b: 24). The manual introduces ‘Betari's Box’ – ‘a model that helps us understand the impact that our own attitudes and behavior have on the attitude and behavior of the people we are communicating with’ (College of Policing, 2016b: 25).

Our attitude plays a large role in the behavior we exhibit. When we are feeling motivated and positive, we have a positive result on those around us. When we're feeling negative, the reverse is often true, we can be impatient, and we appear frustrated or get angry. This in turn means those we are communicating with display those negative behaviors back to us, and the conflict spirals downwards. (College of Policing, 2016b: 25)

The police officer seeks to control the demeanor they project, in order to control the general ambience of the communication. The manual emphasises empathy, listening, understanding, detachment and establishing trust. The police officer is trying to set (and ‘sense’) the ambience. If we read the liaison team trainers’ manual alongside ACPO's communication module training materials (2012), we can see police being actively instructed in affect identification and management. This includes how to avoid an ‘atmosphere of conflict’ (ACPO, 2012: 39), fostering rapport (ACPO, 2012: 9), ‘ego state communication’ (ACPO, 2012: 30) and ‘meta-talk’ (ACPO, 2012: 33). This is well understood in the context of sporting events. The report on the 2011 Stanley Cup riots in Vancouver underlines the importance of mood and atmosphere. In particular it pointed to the role of perky volunteers during the 2010 winter Olympics ‘as setters of mood and atmosphere in the crowd’ (Furlong and Keefe, 2011: 117) – a practice that was also deployed in London's summer Olympics of 2012. However, this ‘tone setting’ is only the beginning of the affective loop.

Unlike forward intelligence officers, the liaison team is not directly tasked with collecting information that might lead to arrests (College of Policing, 2015: 17–19). Instead they are involved in a different type of strategic intelligence gathering. In Adapting to Protest, Her Majesty's Inspectorate of Policing makes clear that the role of the Swedish ‘dialogue officer’, on which the protest liaison officer is based, was to ‘read moods and preparedness for action in the group of demonstrators and how that is affected by police activities and to inform commanders of consequences of different courses of actions in a short and long term perspective’ (2009: 75). When the findings of Adapting to Protest were finally incorporated into training materials, this task of ‘reading moods’ was framed as ‘sensing’: ‘As the [liaison teams] are within or close to the protest crowd, they are better able to sense the mood of the crowd and to assess how police actions are perceived’ (College of Policing, 2016b: 20). The manual insists upon building a personal relationship with various protest groups specifically to improve the quality of this ‘sensing’: ‘As [liaison teams] will often have had long term contact with protest groups, they will be better able to interpret the mood and conduct of the group than someone who is unfamiliar with the group’ (College of Policing, 2016b: 20). If they know the group's dynamics, they hope to be better able to identify when more radical elements begin to hold sway, or when anger, aggression, exasperation or other ‘negative’ moods gather.

The liaison team is initially tasked with setting the tone and sensing when a mood changes in a group. However, their role is then twofold. Firstly, they should feed this information back to the command room. Protest liaison officers are highly specialised, the senior officer in each group will be of a rank of Sergeant, and they are deployed by a Silver Commander and report to a Bronze Commander (College of Policing, 2016b: 15). Thus, they can be granted a level of tactical and operational independence. Equally, they can feed into the decision-making process at a higher level than the ordinary forward intelligence officers. They must relate the group mood and the atmospheric changes to their disorder model and knowledge of crowd psychology and dynamics. This complex picture is then relayed to the command structure through ‘fast time’ updates, allowing the control room to add the affective terrain to their real-time surveillance of the situation. Secondly, the protest liaison team should try to de-escalate the situation, to ameliorate any ‘negative’ mood changes. The trainers’ manual explains this: the team should ‘interpret and explain the actions of the police to organisers [of the group/protest] in an attempt to prevent negative responses from the protest crowd’ (College of Policing, 2016b: 20).

The liaison team and the protestors create the first part of the affective feedback loop. The trusting atmosphere between them allows the circulation of information which can encourage the protestors to self-police in advance, or de-escalate tense situations on the day. The loop then continues through the police command-centre, with the liaison team's fast-time updates allowing them to develop a complex picture about the affective and atmospheric effects of particular interventions. While the quality of these updates is always going to depend on the ability of the protest liaison officer to ‘sense’ the mood, it remains a significant advance on the surveillance footage upon which the control room would otherwise rely upon for gauging the effects of their actions. The protest liaison teams allow the atmospheric impact of police action to be gauged in real time against the group ‘mood’ and their ‘preparedness for action’. This is the key element of the ‘strategic facilitation’ model.

Affective Public Order

I do not think that it is overstating the case to suggest that there are affective considerations in all forms of public order tactics: in the selection of police formations, in baton charges, dog handling, horse charges, tear gas or vehicle techniques. Affect management is present in the design of clothing, hats, gloves, shields, vehicles, gas grenades and other public order technologies. One need only think of the Defence Technologies Ltd. gas/stun grenade launchers which are produced in bright toy-like yellow, orange or green for the public order market (2018: 190), or the introduction of the characteristic ‘hiss’ into CS gas grenades to increase their ‘fear-inducing’ character (Feigenbaum, 2017: 66). The affective and atmospheric dynamics of crowds have been a key site of investigation for the police, at least since Tarde and Le Bon's initiation of a systematic analysis of contaigous communication (Schweingruber, 2000). This affect management will only become more important, as ‘mood sensing’ machine-learning technologies are developed beyond their current infancy (Dartnell, 2018). Affect management is a crucial element of public order policing, yet it is rarely considered in these terms. Instead you find discussions of the symbolic significance of police action or their excessive use of force. An affective reading of these techniques can help us deepen our understanding of the ‘forces’ used by the police, and open onto a broader set of questions for protestors.

This article has identified some of the key atmotechnic interventions that police have developed to change the comportment of crowds. What remains to be noted, however, is just how important their attempts at atmospheric change can be. We have seen that atmosphere is the affective tone of a space, a tone that can be changed as the space is performed in a different way. However, what I have not mentioned is that atmospheres tend to be experienced most intensely as they shift. We might imagine a nightclub towards the end of an evening: the warm moist air, the dim lighting, the smell of bodies, the pounding music. But then, with a certain inevitability, the music comes to an abrupt halt and the house lights come on. There is a shock as one suddenly moves from one atmosphere to another. ‘At the point of becoming-other … an atmosphere becomes visible, ontologically vibrant. This vibrance becomes epistemologically accessible as … [a] different atmosphere’ (Philippopoulos-Mihalopoulos, 2015: 165). The ‘vibrance’ is a sudden bodily attention to the atmospheric shift. Comportment changes – people stop dancing and begin to gravitate towards cloakrooms or the exit. Brennan is clear that we usually sync with the affective dynamics of a situation – a process called entrainment (2004: 68). Thus, we might say that this ‘vibrance’ is the experience of atmospheric shift. The atmotechnics of fright and might that we saw in the ‘escalated force’ model are aimed at precisely this moment of vibrance. By surprising protestors, the aim is to suddenly intervene in the atmosphere, encouraging them to scatter through the vibrance of the atmosphere. The problem for the police, however, is that if the crowd does not disperse, the show-of-force has established a highly charged and aggressive atmosphere. This is more likely to precipitate violence (CCPOST, 2012: 31). This too is entrainment, as the police and protestors normalise to a new aggressive atmosphere, both with a growing sense of their own strength.

While ‘escalated force’ is no longer the primary strategic model for dealing with protests, the specific tactics that formed the model remain a central part of police training. The 2004 manual talks of police deploying a ‘show of strength’ or explains how they ‘register a presence’ (ACPO, 2004: 124); the 2016 College of Policing's National Police Public Order Training Curriculum extensively discusses the ‘show of strength’ to be deployed before the use of batons (2016a: 17–19); module E1 specifies how dogs can be deployed to increase the affective impact of a ‘show of force’ (College of Policing, 2016c: 8–14). The remnants of ‘escalated force’ are present in each successive public order ‘model’. Body-Gendrot points out that the police ‘oscillate’ between the different models in response to any given protest (2017: 97). This means that ordinary police officers must be au fait with all but the most problematic of techniques. The latest public order model is significant, however, as it actively seeks to set a different tone for protests. It is no longer a matter of shocking protestors with the potential violence of the police, of negotiating with leaders, incapacitating ‘trouble-makers’ or ‘containing’ unruly groups. The police seek to build relationships with protestors in order to more efficiently and effectively deploy their affective force. The feedback loop is perhaps the most important but as yet unremarked upon development. It means that the police control room can test tactics that often get the greatest resistance, while understanding the effect that interventions have. In short, the latest model contains a dynamic assessment of protest mood and a sense of the protestors’ capacities to act. This advances the police significantly beyond the original linear approach to collective mood and atmosphere. However, it is best seen as a meta-tactic, allowing police an insight into collective moods, rather than replacing older tactics.

Conclusion

It is worthwhile to return to the start of this article and insist upon the importance of the atmospheric techniques themselves. Without ever claiming that these techniques necessarily work, or work as they are imagined, it is possible to begin to gather an atmospheric praxis of protest. The last 35 years of manuals and training materials that make up the police archive underline an increasingly nuanced understanding of the force of affective techniques. But as well as being a rich vein of analysis for affect theorists, the archive also contains a fundamental challenge for theorists (and tacticians) of protest. Just as police seek to gain control over bodies and space through atmotechnic interventions, protestors need to develop techniques that immunise crowds or disrupt the affective forces that are brought to bear upon them. The police archive suggests the possibility of a counter-archive of protest atmotechnics. Such a counter-archive would have to entwine the analyses of objects, spaces and behaviours with the more ineffable and transient moments of intensity and exhaustion. This counter-archive would not re-articulate the familiar repertoires of protest (Tilly, 2006) but sensitise us to the atmospheric dynamics of those repertoires. In other words it would begin to develop an atmotechnics of protest, focusing on those techniques that generate particular atmopsheres that are conducive to protest.