Public Positions : Police against the People

Monday, 14 April, 2014 - 16:30

At 4:30pm on Monday April 14 the second event in the Public Positions series will be hosted at WISER.

Streaming for this event, from 4:45pm, is available at

A short paper produced by Julia Hornberger will form the basis for a lively discussion by a panel chaired by Firoz Cachalia, and consisting of Kelly Gillespie, Gareth Newham and S'bu Zikode.

In this session we examine the crises around popular protest and policing.   Why is there so much violence associated with policing protest? Is this a matter of policy and training? Is the relationship between the people and the police formed by our history? Or is the police force carrying the burden of the entire state, for many of the poor the only instance of the bureaucracy available at the local level? Recent events at Relela and elsewhere suggest that this problem – frighteningly interlaced with the limits on the state and popular convictions about witchcraft – threatens to unravel the fabric of citizenship. These are all problems that have been richly examined in South African social science over the last century.

We need a complicit police!  Political policing then and now

By Julia Hornberger

The demise of public order policing

On 13th of April 2011 Andries Tatane, a teacher and local activist, was killed by officers of the South African Police Service (SAPS) during a protest in Ficksburg, a small town in the Free State, South Africa. Although the killing had been captured live on video, all seven police officers involved in the incident were acquitted of the killing as the state failed to prove beyond reasonable doubt who exactly had fired the deadly rubber bullets, and that there was ‘common purpose’ by the police officers involved in killing Tatane. For many this tragic and brutal killing is seen as a watershed moment, marking the definite return of police violence (repressive violence) well known from apartheid times (Bruce 2012a, 2012b, 2013; Jacobs and Wasserman 2011, Chance 2013). It brought broad public awareness – not least because the event had been captured on video - to a chain of similar incidences of police brutality during protests both prior to and after that of Tatane including, of course, the Marikana shootings. These amount to a frightening picture of police failure to deal with public protests in a democratically acceptable manner.

One irony of the moment, however, is that in the last 15 years the SAPS has been substantially disinvesting itself from public order policing in order to deal with the pressure to reduce crime. In an effort to increase its legitimacy with the people of South Africa and its administrative accountability it has been putting the bulk of its - both budgetary and personnel - into ‘ordinary’ crime fighting. In 2000 the Mbeki government released the National Crime Combating Strategy (NCCS), a belated reaction of a government that was accused of not taking crime seriously. Soon after, in 2002, began the dismantling of the Public Order Unit, which by 2007 was reduced by more than 64% in terms of personnel. Officers made redundant were relocated to bolster everyday police station work. And those remaining in the renamed Crime Combating Units, while nominally still responsible for public order, were also in practice re-deployed more or less full time to major crime fighting operations (Omar 2007).

This left the public order police understaffed and undertrained, and unable to deal with the public dissatisfaction with a failing government. Yet, in the light of history this trend could be considered an unprecedented and even a progressive move. Policing in South Africa has generally been characterised as being overused for crowd control purposes and underused for crime fighting purposes. And even where the police concentrated on crime fighting, these were often crimes that served as yet another way to keep black South Africans in a subjected state (e.g. Pass Laws). This bias was accentuated by the fact that police was thinly spread, forcing difficult choices as to where to deploy its resources (Killingray 1986).

Political policing then

From inception, one of the primary roles of the police, mounted and in paramilitary fashion, was not to keep peace among people but to police territory and suppress internal resistance to colonial rule (Brewer 1994). These colonial regiments - mounted riflemen - at least in the British territories of Natal and the Cape, followed the model of The Royal Irish Constabulary that had a long history and proven record of suppressing civil unrest and political agitation (Killingray 1986).

Prior to Union in 1910, there was also not just one police force. Mounted regiments were complemented - and that in at least quadruplicate, for each of the colonies - with a potpourri of other police forces, such as special police for key infrastructures, e.g. the railway police, private police for the mines, native administration police and town police. But even where there was town police, like the one set up in Johannesburg at the turn of the previous century, which supposedly subscribed to a more civilian outlook, this was put very much in service of supporting the mining industries in forcefully managing its workforce. Three laws deserve particular mention here: The Liquor Law, the Gold Law and the Pass Law. Together, their enforcement led to the mass incarceration of an otherwise innocent black population (Breckenridge 2014).

With the forging of the Union, the plan was to have one single and highly centralized police force, or at least that was the fantasy of newly appointed police Commissioner Truter (Dippenaar 1988). Truter succeeded in centralizing the force with control located firmly in Pretoria. (This is a peculiar but consequential development, which still pertains; it means that police stations and all other units are primarily accountable to police headquarters rather than municipalities (Breckenridge 2014). It is very different for example from what happened in other colonies such as India, where local elites collaborated with the police to define what constituted criminal activity (Singha unpublished). Although decentralisation could potentially exacerbate local despotism, it also meant that police was much more accountable to the locality in which they were operating).

However, regarding the second aspect of having only one single police force the government (particularly the Ministry of Justice) insisted on keeping a dual system: the South African Police (SAP) for the burgeoning cities, and the South African Mounted Riflemen (SAMR) for the countryside and the control of ‘tribal rivalry’ and resistance to white rule. To leave no doubt about the role and method of the SAMR, it was promulgated under the Defence Act of 1912 instead of the Police Act of 1913. The SAMR was absorbed into the SAP finally after the Second World War. By then, however, unrest had become a phenomenon of urban areas much more than of rural areas anyway, and the SAP had already taken over many internal security tasks, such as the quelling of protest and strikes. In fact, from its inception in 1913 the SAP was fully absorbed by such tasks (Brewer 1994). Take a year such as 1914, in which the SAP first helped to suppress a railway strike, which turned into a general strike by white workers, and was swiftly crushed by the police and military acting with powers under martial law. Then, in the same year, the police were involved in suppressing the de la Rey rebellion, and finally the police helped with the conquest and occupation of German South West Africa. Each of these interventions involved the killing of people and often left those being crushed ever more united and politicised than before the intervention. On the side of the police it drove a process of militarisation, with a bias towards drill and weapon training and the introduction of military ranks in 1919. By 1922 it had even become thinkable to use the police in combination with airforce bombing to end a strike by white workers (Killingray 1984). Still the number of people killed in those interventions paled when compared with how the police dealt with black resistance. In 1920 the SAP, lead by the Commissioner himself, killed 200 black people in an uprising in Bulhoek (Brewer 1994: 100).

Meanwhile, where the police was trying to deal with so-called ordinary crime, which threatened white people’s lives and property in the growing and industrialising cities, and which was rife especially on the Witwatersrand, it was deeply caught up in inefficiency. This was shaped especially by corrupt entanglements with the various gangs and gangsters who –attracted by the unruly, male dominated capitalist precious metal business of early Johannesburg– populated the Reef (van Onselen 2009). Also, the main efforts of the Criminal Investigation Department (CID) still went towards the disciplining of a black mining workforce and the enforcement of the gold, liquor and pass laws. In fact, the ongoing raids of mining compounds and black living quarters in the city in the name of liquor law formed an everyday version of crowd control. Yet, implausible as it might seem, police management tried to maintain a language of modernization and an aspiration towards professionalism and independence by trying to get better educated recruits, by insisting on a civil police spirit, through technological advances in the field of forensics, and by holding up a principle of the use of minimum force. Yet these efforts only appeared in pure form in the wishful language of commemorative albums (Dieppenaar 1988) and the recommendations of various Commissions of Inquiry (e.g. the Commission of Inquiry into the Witwatersrand Riots (1913), Te Water Commission (1926)). And while they might explain why a purely instrumental understanding of the police, as the agent of dominant interests, does fall short, it is often the police itself which failed to concede to itself its highly compromised character and its fundamental role in political policing. This is evident in budgetary priorities. Brewer shows that “The proportion of the police budget spent on detectives services, a measure of expenditure on civil police work, had fallen by a third in the 1926/7 financial year compared to the 1914/15 financial year, while overall the police budget had doubled” (1994: 67).

And even where white citizens, who might have had some influence on what kind of police they wanted, expressed their unease about armed police officers patrolling their area, such liberal concerns were quickly overruled – with the very consensus of these citizens – when confronted with a growing black urban under- and working class (Brewer 1994: 91).

This structural constellation of bias toward crowd control in its exceptional form – mainly directed at white industrial strike action and Afrikaner rebellion ­­– as well as in its mundane form – mainly directed at a black working force – was from its inception at the root of the SAP. It reproduced itself over the years in different variations of course, with different strength and proportions. To mention one more important in this long history: In 1976 the Soweto protest erupted. Like today, riot policing as it was called then had been in a kind of slumber. Political resistance had been quelled in the early sixties and the previous years had been relatively quiet in terms public protest. When the protest happened police intervened brutally. This happened partly out of its ongoing political policing mission. But it also happened out of sheer incompetence and unpreparedness (Jeffery 1991). The reaction of the police in the following months and years was to deal with this unpreparedness by increasing its riot police manpower, to strengthen its chains of command and protective measures for police officers (Brewer 1994). This culminated in the highly militarised police of the 1980s, with very little capacity and will for ordinary crime control.

The desire for a strong state

We have to remember, of course, that the political bias is exactly the bias that left black areas to their own devices in terms of creating provisional means of safe living.1 This gap was filled at different times with different formations of informal justice. It was a form of self-rule which sometimes was politically (locally) legitimate, sometimes highly divisive in inter-generational and class conflict; it often got out of hand and turned from sanctioned force to menace; sometimes it was initiated and even paid (or better, underpaid) by the state, sometimes it was reigned in by the state. Mostly, though, it was just ignored as a necessary if not useful evil in a divide and rule policy (Baker 2008, Buur 2006, Buur and Jensen 2004, Glaser 2008, 2000, Kynoch 2005, Jensen 2005, Kirsch and Grätz 2010, Abrahams 1998). It normalized an experience of a lack of security as public good and of highly authoritarian and rather immediate forms of punishment (Gillespie 2014). Together with post-apartheid’s democratic promises of inclusion and a new infrastructure, such as community policing, which has brought the police closer to the people (Hornberger 2011, 2013), this has produced a highly ambiguous yearning for the force of the state — a yearning which often takes the form of a mainly private relationship with the state (White 2013). The expectation is that policing intervenes forcefully (not particularly hemmed in by human rights) in one’s own favour and protection (Hornberger 2011).

This comes through for example in the policing of domestic violence where the call for the police is often an expression of the desire for a protective but authoritative figure, who can at least match the husband’s violence and reign him in on behalf of the woman. But it also comes through in the policing of public protest. In fact, I would like to propose here, to put matters starkly, that public order policing is not very different from and just as protracted as the policing of domestic violence. The violence itself is the result of a failure of communication and symptomatic of conditions of (gender) inequalities and economic disempowerment. Most importantly it is the epiphenomenon of a structural situation which the police alone cannot change. Like in many domestic violence cases then, most women do not want to get rid of their husbands as such,2 but simply want them to behave differently. We could say the same about municipalities. It is mostly not the legitimacy of the government as such which is at stake but rather how things work or not which leads to protest. Still it is often the police which is called upon. In the case of domestic violence, the intervention of the police is often desired with the hope that it will change the behaviour of the husband at least in the here and now, as the fight is happening (Altbeker 2005, Hornberger 2009, Steinberg 2008). When it comes to public protest, the call is sometimes directly on the police to deal with a particular case that is creating great insecurity within a community. The call may in fact be on other parts of the local state to deliver, but the police remain the most tangible visible manifestation of the state, and becomes the addressee to receive the message (of anger)3. The issue of policing thus serves as a rallying point to hold government accountable and make the suffering heard. This is an important point, and marks a huge difference compared with the late apartheid era: there is a demand for actual policing.

Yet the manner in which this demand is responded to is what leads to a constant disappointment and instead produces a major antagonism towards the police. As with domestic violence, when the police intervene in public protest, they often appear to be intervening on behalf of someone else (Holdt et al. 2011). I reserve my judgment at this point if this someone else is a real or imagined someone else. There is definitely a spectrum of possibilities between a police that is being ordered politically to crush a protest with well known apartheid policing methods, and a police which acts ‘merely’ with a bias towards its own occupational rationale of self-defense and preservation of authority, but with such incompetence that it translates into a policing-against-the-people.

Either way, it is then that the intervention of the police mostly disappoints and in fact aggravates the situation. It is not only that the police cannot solve the situation, but that the very act of policing produces retaliation. Protesters might feel violated, shut up, and the sense of violence suffered is recast as a form of political sacrifice (Chance 2013), meaning that people go into subsequent protests with the expectation that further sacrifice might be necessary. If nothing else gets solved, it is most certainly the case that the police intervention will only produce more need for intervention.

Making things worse

This ability of the police to aggravate the situation is often highly underestimated and misunderstood by those who order the police to intervene, even more so by the police themselves (de la Porta 1998). There might be some theoretical awareness that the police can choose between a calming or escalatory approach, between a minimum of force and a maximum of force approach (Mthethwa 2014a). And to be fair, the police have been retrained in public order policing and this has actually be seen as one of the successes of transformation (Marks 2005), at least before the public order police was dismantled. But the police are still being used as if they are outside of the conflict itself, and as if they can be used as a surgical instrument which can stitch the situation, by removing the trouble or quelling the spilling of blood. They are not seen sufficiently as an integral element of the conflict, though.

A recent ethnographic description of protest against the hosting of the World Cup in Brazil remarks how quickly protesters’ sentiments regarding the police can turn. When the protest started, people mixed their anti-FIFA messages with the message of “sem violencia “ (without violence) hoping for a pact of solidarity with the police. But the police did not respond to the call and instead used a pre-emptive display of might and violence, occasionally throwing stun grenades and preventing the demonstrators from moving to the centre of town to deliver their message. While the demonstration did not turn violent, the author powerfully describes a sense of disenchantment about what is politically possible: “While the chant “without violence” didn’t lose its poetry, it didn’t move me the way it did before the interruption. The pact seemed to have been broken” (Durão 2013).

So-called crowd psychology has provided substantial insights into these sometimes very subtle but potentially highly consequential dynamics, described here from an ethnographic perspective. It shows how police intervention is absolutely crucial to what happens at a gathering; how police in fact have the possibility to either give people the sense that movement is possible or rather that a horizon is closed; and how police are mainly responsible themselves in creating an escalating hostility. An important point is that crowds are hardly homogenous. There is always a broad spectrum of people available in a crowd, from people who are quite willing to self-police themselves to people who are more prepared to use violence (Reichert 2007, von Holdt 2011). It is through police action, which confuses the acts of a few with the acts of a whole crowd, and which creates the crowd as a homogeneous (and violent) entity, which leads to a crowd’s uniting and halting communication. This has been definitely the case in Marikana, where the criminal acts of some tainted the whole group of demonstrators as criminal. To avoid an escalation of violence - in the moment as much as over a long period of time – the police would need to always assume that the crowd is there to deliver a message and that the primary role of the police would be to facilitate the deliverance of the message. The police cannot change inequality and unemployment, but it can choose the side of the protestors and help to deliver the message. Here is a good example of how this should go, taken from Waddington (2007) who was observing the negotiations between representatives of a far left anarchist group and Metropolitan Police officers during the early 1990s in Britain. The declared aspiration of the protesters was to ‘tear down the fabric of capitalism’, to which the Superintendent conducting the meeting replied, ‘And how can we help you?’

Quo Vadis political policing?

However little of such radical mind shift in the approach to police intervention seems to be on the horizon. Instead, like after 1976 today’s primary reaction to the failure of the police to deal with public protest has been the promise to bring back and even build more public order capacity than ever before. Admittedly under a paradigm of the rule of law4, but soon there are supposed to be 9000 police officers ready to deal with community protests all over the country (Mthethwa 2014b). It should be clear that as nothing much might change in terms of people’s demands, the role of the police is pretty much set to become an ongoing occupying army. Unless, maybe, they rather stick to their course of getting ordinary crime fighting right, and to at least make sure that protest does not increase because of their interventions. And so they better learn how to make protest work. What we need is a complicit police, complicit not in the inertia, but complicit in making things move.

1 The same kind of analysis is not appreciable for apartheid rule of coloured areas, where police played a much more crime control oriented role, but in the process subjected coloured families, especially coloured men to a carceral/reformatory regime (Jensen 2009).

2 As this is a paper about public order policing, I wont go into too much detail with regard to the complexities of domestic violence and that it of course happens not only in the relationship of married heterosexual couples.

3 This idea of development by proxy, we can also see with regard to the recently held Khayelitsha Commission. The situation in Khayelitsha is hardly just about policing but about a lack of development and a lack of work.

4 Note that according to a strict application of the rule of law most protest would be illegal as they rarely can get the approval of the municipality or councilor.

WISER Research Theme: