South African journalism and the Marikana massacre: A case study of an editorial failure

Jane Duncan, Rhodes University

Keywords: Marikana massacre; South African media; political economy; workers in the media; investigative journalism


This article examines the early press coverage of the police massacre of striking mineworkers in Marikana, South Africa, on the 16th of August 2012. It includes a quantitative and qualitative content analysis of a representative coverage sample from the country’s mainstream newspapers. From this analysis, it is apparent that the coverage was heavily biased towards official accounts of the massacre, and that it overwhelmingly favoured business sources of news and analysis. Business sources were the most likely to be primary definers of news stories. In contrast, the miners’ voices barely featured independently of the main trade union protagonists, which was significant as many miners did not feel sufficiently represented by the unions. The failure of journalists to speak to miners sufficiently led to a major editorial failure in the early press coverage, as it failed to reveal the full extent of police violence against the miners. As a result, the police version of events was allowed to stand largely unchallenged before a group of academics put the fuller account into the public domain after having interviewed miners. The coverage also contained dominant themes which portrayed the miners as inherently violent, disposed to irrationality and even criminality. The article traces the problems that arose in the coverage, specifically, in regard to organisational and occupational demands in South African newsrooms. More generally, problems in the coverage also reflect the transition of South Africas political economy. This transition has reproduced and reinforced the countrys social inequalities. These circumstances have facilitated the growth of police violence against workers and the poor.

On the 16th of August 2012, 34 striking mineworkers were killed and 78 were injured when members of the South African Police Service (SAPS) opened fire on them in Marikana, a mining town in the North West province of South Africa. The massacre took place in the context of a strike over pay at a mine owned by British multinational mining company Lonmin PLC. The mineworkers shunned the formal procedures for protected strikes set out in the country’s Labour Relations Act. They also shunned representation by the majority union, the Congress of South African Trade Unions (Cosatu)-aligned National Union of Mineworkers (NUM), which was an integral part of the formal bargaining framework, and engaged on an unprotected strike. This they did because they felt let down by the formal bargaining system. Their key demand was for a salary increase of R12,500 (USD1,211). Lonmin maintained that this demand was unreasonable and would impact negatively on the company’s viability. Ten people were killed before the massacre, between 11 and 15 August, including security guards and policemen. The media reported extensively on these events, and continued to report on their aftermath [1].

In the wake of the massacre, President Jacob Zuma, head of the African National Congress (ANC)-led government, appointed a Commission of Enquiry into the massacre, headed by Justice Ian Farlam. The Commission began its work on the 1 October 2012. Initially intended to be a four month affair, the duration of the Commission was extended by the President. The police maintained that they acted in self-defence as armed miners attacked them first. As the Commission’s work continues, this narrative is starting to unravel. Emerging evidence suggests that the violence against the miners was more premeditated than the police have been willing to concede.

Police violence is a systemic problem in South Africa; it is not just a case of a few rotten apples. In a number of recent cases, SAPS members charged with the killings of civilians, including protestors, have been found not guilty for lack of evidence. Furthermore, the National Prosecuting Authority has developed a reputation for selective and politically partial prosecutions to favour the ANC’s current leadership. These developments strongly suggest that the criminal justice system is not serving ordinary South Africans as it should, with the result that police violence could well continue, and even escalate. An unavoidable question presents itself even before Farlam completes his work: what if these processes fail to deliver justice?Undoubtedly, many want journalists to continue the search for the truth. But will journalists be up to the task of holding the police to account when failing institutions are unable or unwilling to do so? More specifically in relation to Marikana, two crucially important questions present themselves: what happened and who is responsible? Many will look to journalists to answer these questions if the criminal justice system fails to do so.

This article looks at whether journalists are likely to rise to these challenges, based on lessons that can be drawn from the press coverage of the Marikana massacre. Coverage of this event has become a litmus test for how social conflict is represented in South African media.

An overview of South Africa’s political economy

South Africa’s mineral wealth is massive. The country holds the world’s largest platinum reserves, and Marikana lies in the heart of the country’s platinum belt. Over the past few decades, South Africa’s economy has become heavily reliant on the mining industry for economic activity, and by the 1980s, the economy was dominated by six interlocked mining and finance houses (McKenzie and Pons-Vignon, 2012: 6). This economic configuration has become known as the minerals-energy complex. It makes South Africa overly dependent on the export of the country’s rich mineral deposits, mined by labour drawn from an extensive migrant labour system. Profits are beneficiated elsewhere, at the expense of local industry (Ashman and Fine, 2013).

South Africa’s transition to democracy in 1994 was one of a series of negotiated settlements that took place at that time. By that stage, Soviet-style communism had collapsed, the international balance of forces had swung against liberation movements generally, and armed struggles for revolutionary change had waned. The neoliberal phase of capitalism was entrenched and had become practically the ‘only game in town’ ideologically. While the ANC attained political power through negotiations, parallel negotiations with the then captains of industry ensured that the basic structure of the economy remained unchanged from apartheid. Its commanding heights remained largely in white hands (Terreblanche, 2002). This truncated transition has meant that while the formal trappings of apartheid have been dismantled, the social relations forged under apartheid remain largely intact. This has placed significant constraints on transformation at all levels of the social formation, including the media.

South Africa’s path to neoliberalism did not involve brute force, as it did in Chile or Argentina; neither was it domesticated largely by consent, as was the case in the United States and the United Kingdom (Harvey, 2005: 39–63). Nor was it imposed as part of a package of conditionalities demanded by the International Monetary Fund or the World Bank, in return for loans. Instead, neoliberalism was self-imposed in the late 1990s, shortly after the transition to democracy, through the Growth, Employment and Redistribution Plan (Gear). The plan included a package of reforms designed to make the economy globally competitive by increasing its external orientation. This was coupled with drastic cuts to public expenditure and the privatisation of strategic state assets. Developed in response to a speculative attack on the South African currency, the Finance Ministry quickly declared the plan ‘non-negotiable’. In the early 2000s, the government began to roll out an extensive network of social grants and other social wage top-ups, such as free basic water and electricity allocations, to soften the immiserating effects of neoliberalism and ensure at least some level of consent for this shift.

However, far from stabilisingthe economy, the imposition of Gear heightened its financialisation, at the expense of the productive economy. Unemployment levels increased, especially amongst youth, many of whom faced a lifetime of joblessness as part of a growing surplus population. While poverty levels decreased, inequality increased. In an attempt to cut public spending, local governments were forced into financial self-sufficiency; yet, at the same time, they were being expected to extend services like water, electricity and sanitation. Poorer municipalities with smaller revenue bases could not square this circle very easily. There were massive service delivery problems, and even an outright collapse of services in extreme cases. Gear also led to massive capital flight from the country, as South Africa’s largest companies listed overseas these capital outflows were not matched by an increase in foreign investment. As a result, the current account deficit ballooned. The foreign exchange inflows that did take place were largely speculative in nature.

These problems made attempts at securing consent for the prevailing socio-economic order unsustainable over a longer period. The superficial consent that had been constructed was likely to unravel. In fact, signs of growing class polarisation emerged in the second decade of ANC rule when large numbers of voters abstained from voting. The ANC still enjoyed a healthy majority, but off a declining voter base. Grassroots resistance to the effects of neoliberalism began to erupt all over the country; however, protests were sporadic and have not coalesced into a national movement with an explicitly anti-neoliberal politics. Nevertheless, there are signs of greater political diversity emerging.

Industrial struggles have ebbed and flowed, mainly because the main trade union federation in the country, Cosatu, maintained an alliance with the ANC (arguing that the alliance increased its influence over government). However, the arrangement has compromised the ability of the movement to act as an independent voice of class struggle in the country. Until recently, its largest affiliate was NUM. However, workers became increasingly disaffected with its ability to represent them effectively, leading to many joining another union, the Association of Mineworkers and Construction Union (AMCU). This sparked a deadly turf war between the two unions. Other workers began to establish independent workers’ committees to negotiate on their behalf, outside of the formalised collective bargaining structures. South Africa continues to rely on cheap, black labour for its mining sector, and the 2008 global recession intensified financial pressure on workers. Many fell victim to the relatively unregulated micro-lending industry to finance basic living expenses. They became trapped in debt spirals. These factors prompted a wave of strikes in the platinum industry throughout 2012. These occurred in the mining operations of Lonmin, which operates mining plants in Marikana, North-West province, and in the Limpopo region, the northernmost province in South Africa.

At the same time that class polarisation was growing, the ANC government was increasing the coercive capacities of the state. From the early 2000s, SAPS underwent a process of militarisation, which intensified shortly after the government’s most recent president, Jacob Zuma, was installed. While the most visible manifestation of police militarisation was the reintroduction of a military ranking system that existed under apartheid, the number of paramilitary policing units grew and incidents of police violence also increased. More facilitative approaches towards protest policing were gradually edged out in favour of more militarised responses, and paramilitary units began to be deployed in protest ‘hotspots’, including Marikana.

The military was also deployed to supplement domestic policing. Both the military and the police were supported by intelligence agencies which began to characterise protestors as criminals and threats to national security and stability. According to the Minister of State Security, Siyabonga Cwele (the Minister responsible for the civilian intelligence service), “…[violent] industrial action tended to be protracted, illegal, unprotected, and disruptive to key sectors of the economy, with a new trend of the shunning of union representation and hard-won established labour relations dispensation in South Africa” (Cwele, 2013: 2). Referring to the strikes in the platinum sector as illegal is problematic. In terms of South Africa’s constitution, workers have a right to strike and this right is constitutionally protected. However, the country’s Labour Relations Act makes a distinction between protected and unprotected strikes [2]. More workers have chosen to go on unprotected strikes, as protected strikes have proved increasingly ineffective in raising workers’ standards of living; in fact, workers have become poorer in absolute terms (Gentle, 2012). Furthermore, Cwele’s statements about the shunning of union representation and labour relations procedures fail to grasp the fact that workers were shunning these organisations and procedures because they were not addressing their grievances. In fact, Cwele’s comments strongly suggest a judgement in favour of NUM: an inappropriate display of partisanship from the intelligence service’s most senior political leader.

South Africa has a complex media landscape, which has changed markedly since apartheid. Undoubtedly, the media are more representative of the society in which they operate now, and also are largely free to broadcast or publish as they see fit as most censorship laws have been repealed. However, these changes have been premised largely on the commercial media model where market forces have been allowed to shape the media system, with limited public funding being provided for public service content. Broadcasting remains the most popular and accessible medium in the country, with public broadcaster, the South African Broadcasting Corporation (SABC), reaching more South Africans than any other media organisation. However, it has proved susceptible to government control and receives limited public funding. The overly market-driven nature of media transformation has meant that the social inequalities of South Africa have shaped the media system. It has developed the character of a funnel, whereby higher income groups have access to a plurality of media and their media options are expanding all the time. However, media access tapers off further down the funnel.

The press are the news agenda-setters in South Africa, which places them in a unique position to ‘speak truth to power’. Much of this has to do with the fact that they have re-invested in quality content, especially investigative journalism, after a period of massive cost-cutting following the financial crisis in the late 1990s. Press ownership has taken on an oligopolistic character, after an initial period of post-apartheid diversification following the transition to democracy. One large group dominates the press—Media 24, followed by four smaller groups (Independent Newspapers, Times Media Limited, Caxton and the Mail and Guardian). The most prominent investigative teams are linked to the small but highly influential Mail and Guardian newspaper, the Sunday Times (the largest circulation newspaper in the country, owned by Times Media Limited) and its competitor, the City Press (owned by Media 24 and also published on a Sunday). The press is highly reliant on advertising as a source of income, relative to the cover price. This ties the press very directly to the structure of wealth in society. Because the distribution of wealth is still highly racialised and gendered, the still largely white and male upper income brackets remain the most highly prized press audiences because they attract advertisers. As a result, the press has a relatively small and elite footprint, although the advent of tabloids has changed this picture as they have extended newspaper readership to more working class audiences. Newspaper circulation is confined mainly to urban areas. Shrinking circulation coupled with declining profits, especially in the wake of the global recession, are placing newspapers under pressure, as are retrenchments of editorial staff. The pressure to cut costs increases the temptation to be risk averse in matters editorial, and to rely increasingly on agency copy and tried and tested formulas. In this regard, the main wholesale news agency that is relied on most is the South African Press Association (SAPA).

Quantitative content analysis of early press coverage

I will now examine how public understanding of the massacre was constructed by the agenda-setting press, and how its construction relates to practices of power in society. To explore this theme, the press reporting during the crucial period leading up to the massacre and in its immediate aftermath was analysed. It was during this period that the media assisted in forming public perceptions about the massacre, including the events that took place, the causes and the blame apportioned to the various actors. In total, 153 newspaper articles over the period of 13–22 August 2012, provided by The News Monitor via Media Tenor, were analysed [3]. Online news articles were not included. The articles analysed were coded for sources in order to establish who the primary definers of the story were. A lexical analysis of key terms associated with the massacre was also undertaken. All the articles that contained the words ‘Marikana’ or ‘Lonmin’ were included. The source analysis included people and organisations who were quoted directly, or who clearly provided information that formed part of the basis of the article. Many articles had several sources. The sources were then grouped into broad categories, including business sources, political parties and Parliament (most parties were quoted speaking in Parliament), independent experts and commentators (that is, independent from any of the other source categories). The other categories were Lonmin mine management and owners, government, the police, the workers and worker organisations. The result of the source analysis is as follows:

Figure 1. Sources of information for press coverage, 13–22 August 2012. Source: compiled by author.

The business voice, which included institutions like the Chamber of Mines (South Africa’s mining industry employer organisation), banks and risk assessment companies, dominated, providing 27% of sources. Mine management and owners followed, which included sources like Lonmin management, Shanduka (Lonmin’s empowerment partner), and the then Executive Chairman of Shanduka, Cyril Ramaphosa. The ‘other’ category included any sources that did not fit into the other categories. Miners’ voices outside of the two main union protagonists NUM and AMCU, accounted for a mere 3% of sources: the lowest of all the source categories. Of these miners, only one miner was quoted speaking about what actually happened during the massacre, and he said the police shot first. Most miners were interviewed in relation to stories alleging that the miners had used ‘muti’ (traditional medicine) to defend themselves against the police’s bullets. The miners’ working and living conditions were featured in some stories.

So in other words, of all 153 articles, only one showed any attempt by a journalist to obtain an account from a worker about their version of events. There is scant evidence of journalists having asked the miners the simplest and most basic of questions, namely ‘what happened?’ Overwhelmingly they relied on official sources of information and on business perspectives for information and analysis.

The articles were then coded and analysed for the sources that were considered as ‘primary definers’ of the story [4]. Primary definers were categorised on the basis of their prominence within a story. A crucial consideration was whether a given source gave a definitive analysis of events that formed the main storyline of the article. Analysis of primary definers is useful in assessing how social power is expressed through the media, especially if one gives differential weightings to certain social actors and their views (Hansen et al., 1998: 109). The findings are as follows:

Figure 2. Primary definers in press coverage, 13­22 August 2012. Source: compiled by author.

This analysis indicates the findings of the source analysis; business voices were overwhelmingly the primary definers of events followed by the Lonmin mine management and owners. This analysis was coded slightly differently from the source analysis. Journalists’ voices, as well as the voices of former ANC Youth League leaders Julius Malema and Floyd Shivambu (who were by then sharply critical of the ANC [5]) were separated out from the ‘other’ category in which they were placed for the source analysis. It is apparent that the workers had an even smaller share of the primary definer press space than they did as sources.

The articles were then coded for vocabulary or lexical choice, as these expose the underlying semantic frames used in defining the event. These semantic frames may in turn point to the opinions, prejudices and even social position of the speaker (Hansen et al., 1998: 109). In characterising events, a distinction has been drawn between nominal and descriptive characterisations. The former concerns itself with value-neutral descriptions of events, while the latter makes judgements about the description. The main focus of this aspect of the analysis was on the choice of terms used to describe the massacre, and whether these choices could be considered nominal or descriptive characterisations (Hansen et al., 1998: 109).

Figure 3. Top terms used to describe the Marikana story, 13­22 August 2012. Source: compiled by author.

The diversity of terms used to describe the massacre was, in itself, telling. The word ‘massacre’ entered the lexicon of terms the day after the killings, but gradually declined in frequency over the period, to be replaced by ‘tragedy’, ‘killings’, ‘bloodbath’, ‘violence’, deadly clashes’, ‘dispute’ and a host of other variations. The word ‘massacre’ carries with it a value judgement about the events on the 16th August, namely that the police used excessive force. There can be no justification, however convincing, for a massacre. The other terms, though, did not apportion blame for the events, or even single out one actor as the main aggressor and another as the main victim. The reduction in frequency of the term ‘massacre’ could possibly be attributed to the controversy that arose around the use of the term. Some commentators argued that the term was not being used appropriately because it reflected a biased judgement on the event.

It is significant that many of the articles that used terms other than ‘massacre’ also referred to the violence that preceded the massacre, and highlighted the killings of the police and security guards. The lexical choices, when combined with this subject matter, conveyed the impression that there was a moral equivalence between the main actors, namely the police and the workers. In other words, there was no adjudication of the claims made by these actors. The reportage substituted what was arguably a more nominal term (‘massacre’) with more descriptive terms. Yet, these terms were themselves laden with inappropriate judgements, particularly the judgement that the level of force used by the police was proportional to the level of force used by the miners, which was clearly not the case. Automatic rifles, in their ability to injure and kill are simply not equivalent to traditional weapons and handguns.

These analyses of sources used, primary definers and lexical choice strongly suggest that the workers were subjected to what George Gerbner has termed ‘symbolic annihilation’ through under-representation or non-representation (Hansen et al., 1998: 111) [6]. When they were represented, they had little agency in shaping the coverage and defining its overall direction. This suggested that the press constructed a public understanding of events that presented the official account of events as being, at best, largely reasonable, and at worst, an example of police panic at the workers’ growing and increasingly violent militancy. The absence of the workers’ accounts meant that no alternative narrative emerged to challenge the dominant perception of events.

Thematic content analysis of the early press coverage of Marikana

A quantitative content analysis can give only part of the picture, as it limits analysis to those parts of the text that can be counted (Berg, 2007: 238–267). An analysis also needs to focus on interpretation, and not just quantification, as the former helps to uncover patterns in how meaning is constructed in media texts, while the latter concerns itself with the frequency of particular units of analysis. While quantitative content analysis is deductive in nature, qualitative analysis is inductive as it allows major themes to emerge from the reading of particular texts (Hansen et al., 1998: 96). I will now explain some of the major themes that arose in the press coverage to see if the trends apparent in the quantitative analysis are also apparent in the qualitative one. The data set is the same as the content analysis, although some online reports are included as well (specifically those of the Mail and Guardian as their online arm has maintained a constant stream of reportage on Marikana, mainly by Kwanele Sosibo). This analysis includes a mix of the business press and the general press as it could be anticipated that the business press would be more partial to themes that supported the business case, while the general press would make space for more diverse themes.

As noted by Benjamin Fogel in his analysis of the media coverage of the Marikana massacre (Fogel, 2013), a key theme in the earliest reporting was of the conflict being one of inter-union rivalry. The main sources emphasising this aspect of the analysis were the police and Lonmin. The Business Day did, though, refer to the inter-union rivalry allegation as ‘suspected inter-union conflict’ (Seccomb, 2012: 1), leaving space open for other interpretations. The Sowetan was much more definitive in its editorial slant. A kicker to one article proclaimed (SAPA, 2012a: 15) that the violence was as a result of ‘rival unions in a turf war’. Journalists also seemed to assume that including union voices was sufficient to ‘cover’ the workers’ voices, even though it became clear that many miners did not feel sufficiently represented by either the NUM or AMCU in the days preceding the massacre. In fact, according to testimony delivered to the Farlam Commission, 52% of the workers who gathered at the koppie were NUM members, 35% belonged to AMCU, and the rest were not unionised (Sosibo, 2013). One exception to this reporting trend was a story by the Mail and Guardian’s Kwanele Sosibo, which appeared the day after the massacre, but which was clearly written before the massacre occurred. The story quoted a workers’ representative called ‘Nzuza’ as saying that the striking workers gathered on a hill near the mine were not assembled under a particular union banner (Sosibo, 2012a).

Controversy surrounds the events that preceded the massacre, and particularly an incident that took place at the NUM offices in Marikana, although there is general consensus that this incident escalated tensions and precipitated violence that led to a number of deaths. According to several sources, workers, many of whom were NUM members (Sacks, 2012), marched on the NUM offices on August 11. Their intentions remain the subject of speculation (Hlongwane, 2013). NUM had alleged that their intentions were violent, and that the occupants of the office acted in self-defence, but several workers have alleged that they intended to inform NUM that they were going to approach Lonmin independently of NUM (Alexander et al., 2012), but that NUM officials attacked the workers first and critically injured two of them (reported allegations of two workers having been killed in the skirmish have been proven false).

Testimony to the Farlam Commission so far suggests that—barring the allegations of two miners having been killed—the workers’ version of events is the more credible (Hlongwane, 2013). In fact it has emerged that the two workers who were injured, Bongani Ngema and Mabuyakhulu (known as ‘Zulu’), were shot in their backs, and therefore could not have been attacking the office, but rather were retreating (Sosibo, 2013). It would appear that this was when the workers escalated their carrying of arms, for the purposes of self-defence. Yet much of the press reporting repeated the fact that miners were heavily armed with traditional weapons, without any context provided on why they had come to bear arms. These events were then followed by the killing of security guards, reportedly in a second march on the NUM offices (Alexander et al., 2012). This strongly suggested a conflict, at least in part, between a union and its own members; yet much of the early reporting on the violence that preceded the massacre portrayed it as inter-union rivalry, which was—in the case of Business Day (Seccomb, 2012: 1) and the South African Press Association (SAPA, 2012a: 15)—an analysis taken directly from police spokespeople. Furthermore, the press reporting failed utterly to explore the sequence of events that led to the killing of the police and security guards, leaving the reader with the impression that the miners had become violent of their own accord.

However, the miners’ actions cannot and should not be whitewashed purely as acts of self-defence. According to police testimony at the Farlam Commission, the two security guards who were overpowered by the miners were mutilated. One man’s face was hacked and his tongue was cut out and a second was burnt beyond recognition (Sosibo, 2012b). A NUM shop steward, Isiah Twala, was also hacked to death (Sosibo, 2012d). These events suggest use of violence that was in excess of self-defence. Events around this time still remain contested though. Workers interviewed alleged that on 13 August, they went back to the NUM offices to remonstrate with NUM officials (Alexander et al., 2012); The Star Business Report referred to an ‘armed mob’ attempting to force its way into the NUM offices (Faku, 2012a: 13). This value-laden language implies that the paper had taken a position on the culpability of those involved. The AMCU, however, has alleged through their counsel that miners were ambushed by the police on 13 August, leading to two miners being killed and resulting in miners retaliating and killing two police officers (Tolsi, 2013). The police have maintained that two of their number were killed when they were attacked (Marinovich, 2012a). Yet the press coverage does not reflect the contested nature of events; for instance, only the police’s version of events was cited in the Business Day the next day, with the article relying entirely on two police sources (Maoto and Marrion, 2012). The killings of the police and security guards appeared to have hardened police attitudes to the striking miners and led them to focus on disarming the workers (SAPA, 2013).

In the early reporting, there was also scant focus on Lonmin management’s role in triggering the conflict. They insisted that workers’ grievances should be channelled through NUM, rather than (Lonmin) agreeing to meet with the mineworkers’ directly (Alexander et al., 2012). Furthermore, there were references to the strike being illegal, and the inclusion of quotations from Lonmin and the Minister of Mineral Resources Susan Shabangu, referring to the miners engaging in an illegal strike and breaking the law (Rees et al., 2012: 17; SAPA, 2012a: 15; Speckman et al., 2012: 17). No attention was paid to the reasons why the miners chose to go on an unprotected strike outside of official channels. These representations created the general impression that the miners were predisposed to illegal conduct, including violence (rather than as reacting to violence, albeit at times with excessive violence themselves).

Another theme of the early reporting, although subsidiary, and which counterbalanced somewhat the impression created of the workers being excessively violent and unreasonable, focused on the failings of the mining industry to offer decent work and their intransigence in confronting this problem. This foil to the official accounts was a report released at the time by the Benchmarks Foundation [7]. They incorporated six mining communities in a participatory study of the impacts of mining companies’ corporate social responsibility programmes. This report found that unless employment practices and the social conditions experienced by the miners were addressed, then “…we can expect mass uprisings, xenophobic attacks and ongoing disruptions to mining operations” (Capel, 2012: 14). This coverage provided important context that located the violence at Lonmin within the structural violence that the workers were subjected to. The report was used by some journalists to question the official claims that not enough was being done to address workers’ grievances (Speckman et al., 2012: 17).

A much more dominant theme of the early reporting, though, especially in the immediate aftermath of the massacre, concerned the impact of the strike on the financial viability of the platinum sector and mining generally. Here, the most visible sources of analysis were business journalists, risk assessment firms, and the specialist banking and financial services group Sasfin. They all expressed concern that violent strikes would harm investor confidence (Vanek, 2012: 17). The inter-union rivalry that apparently had triggered the violence was, in the words of The Times newspaper, “…causing untold damage to the already struggling economy” (Chauke and Strydom, 2012: 1). Reuters and Bloomberg provided analysis of the strike on the Johannesburg Stock Exchange (Reuters and Bloomberg, 2012). A related theme was of the miners’ demands placing the company’s market value in jeopardy (Faku, 2012b: 17). This framing of the conflict was clearly and unashamedly biased towards the business case, further delegitimising the legitimate (albeit legally unprotected) attempt of the miners to secure a living wage (given that the formal bargaining structures had failed them).

Controversies abound as to whether content analyses should include latent as well as manifest content; while the former focuses on analysing data that is physically present, the latter can extend to symbolic content not explicitly articulated, but nonetheless implied by the physical data (Berg, 2007: 242–243). The problem with relying on manifest content only, though, is that crucial meanings implied by data, but not explicitly stated, may be missed. Ideally both should be employed to decode meaning. The reporting of the actual massacre suggested that the police reacted when they were attacked, with The Citizen and The Star Business Report even describing the incident as a ‘shoot-out’ between police and strikers. They preceded this with references to the killing of the police and security guards, although on the surface most articles adopted a neutral stance with respect to culpability (Crime reporter and SAPA, 2012: 3; Mnyandu, 2012: 18). The manifest content suggested neutrality, while the latent content suggested that the police actions were understandable as they needed to meet firepower with firepower. This is in spite of the fact that there is no evidence that many of the miners were armed. One miner was filmed carrying a firearm, and while the police claimed to have retrieved several firearms, subsequent testimony to the Farlam Commission has strongly implied that traditional weapons were planted on the miners by the police to bolster their case that they acted in self-defence (Sosibo, 2012c).

Reinforcing the latent theme of miners being inherently violent, The Citizen offered an analysis quoting labour analyst Loane Sharpe. He argued that strikes were becoming more violent as unions competed for more space, with no mention made of the role of the police in the violence. This implied that unions were ultimately to blame for the violence that led to the deaths, rather than the police (Clarke, 2012). Business Day continued to insist that those involved in the confrontation that led to the massacre were AMCU members (Maoto and Marrion, 2012). Initially it was reported that 18 miners had been killed, but by the 18th of  August it was reported that the death toll was approximately 34. Curiously, not one report questioned why the official death toll nearly doubled. Some reports were more critical of the police conduct, though, including City Press, which pursued a story line that questioned why the police’s own directive on the use of force during protests was ignored. According to this directive, the police were even forbidden from using rubber bullets, never mind live ammunition (Dommisse, 2012: 6).

After the massacre, journalists also pursued a theme alleging that miners attacked the police on the 16th August because they had been made to believe by a sangoma that muti smeared on their bodies had made them invulnerable to bullets, and that this led to the miners charging the police on the 16th August (Shoba and Mahlangu, 2012). This account of events was promoted by the police to support their contention that they fired in self-defence and that the miners attacked. This viewpoint formed part of the police testimony to the Farlam Commission (SAPA, 2012b). In the case of the City Press, the police’s account became the main account framing their coverage. The newspaper quoted a police spokesperson as probably providing the answer to the seeming riddle as to why workers had advanced on the police, namely that “…we were dealing with people who looked possessed, or believed the bullets would not work on them” (Ledwaba, 2012: 4). A subsequent Independent Newspapers report claimed that some miners had blamed the massacre on the killing of a rabbit. Apparently, they had been warned not to kill the rabbit as its presence brought good luck, but they did anyway (Weekend Argus Reporter, 2012). It appears to be a common assumption that miners engaged in rituals and used ‘muti’ before the 16th of August massacre. But the construction that has been placed on these activities is one of workers being turned into mindless automatons by the ‘muti’, lulled by their own superstition into believing that they were invulnerable to police bullets. An alternative, unstated view is that miners engaged in rituals to build mental strength while remaining aware of the risks involved in any confrontation with the police.

The thematic content analysis suggests that the symbolic annihilation of miners that became apparent through the quantitative analysis, allowed dominant narratives to be constructed that, at best, did little to unsettle the official claims that the police acted in self-defence, and at worst, implicitly favoured the official account of the massacre. These dominant narratives also constructed the miners as irrational and as being prone to making unreasonable claims that damaged the economy. They were also seen as predisposed to violence and superstitious beliefs, strongly suggesting a latent trope of the miners as primitive and even sub-human. These narratives cut across the business and general press, and were implicitly biased towards the police and business cases. Their cases were made to appear more reasonable than the workers’ case, although the notable exception in this regard was the Mail and Guardian. There was an alternative narrative theme that placed the miners’ actions in a more appropriate context, but this was a subsidiary theme to the dominant narrative.

The effect of workers’ ‘symbolic annihilation’ on press reporting

Why is the ‘symbolic annihilation’ of the Marikana workers relevant? The public’s initial view of what happened on that fateful day was shaped by the television footage, shot behind the police line, which showed armed miners rushing towards the police, and which gave credence to police accounts that they were attacked first and acted in self-defence. Yet it has emerged that of the 34 miners killed on that day, no more than a dozen were killed at this site (Marinovich, 2012a). Print journalists also reported from behind the police line, and their accounts largely reinforced the police accounts (Laganparsad and Sutherland, 2012). But a few days after the massacre, the University of Johannesburg’s (UJ) Professor Peter Alexander and his team went out and conducted interviews and found that many miners who were present had a very different account of events. Those interviewed spoke about a second ‘kill site’ where miners were allegedly killed in a far more premeditated fashion by the police. Journalists were not present at this site. They also clearly did not conduct an inspection of the entire massacre site either; had they done so they would have quickly realised that there was a second site, several hundred metres away from the first one. The existence of this second site was readily apparent from the many forensic markings. It was here that eyewitnesses alleged that miners were chased by the police and shot at point blank range, and others who were lying down and pretending to be dead were crushed by police vehicles (Marinovich, 2012b). The forensic markings at the scene were consistent with this account (Marinovich, 2012a).

Because the press had not asked the miners themselves for their version of events, they missed entirely the alternative account, which suggested a much more premeditated attack on the miners, ostensibly in revenge for the killing of two police officers earlier on in the week. Admittedly, reporting conditions at Marikana were extremely difficult. Before the massacre, miners appeared reluctant to speak outside their designated lines of communication, leading to their communications with the media being tightly controlled (Moodie, 2012a, 2012b; Sosibo, 2012a). According to the City Press’s Lucas Ledwaba, the striking mineworkers were operating in a climate of fear and suspicion, especially towards those who were not one of them, including journalists. The mineworkers also viewed journalists as being “…in a way, an extension of officialdom and law enforcement…[Although] through patience and persuasion we eventually managed to gain access to them, this did not necessarily make them trust us or our intentions” [8]. What also did not help matters was that two policemen who filmed the mineworkers masqueraded as journalists until miners identified and confronted them. The miners did, however, expect journalists to be more sympathetic to their plight, and did not take kindly to any questions about their role in the violence that led up to the massacre [9]. On the day of the massacre, the police declared the area where the miners were gathered an operational area, forcing journalists to stay behind the police lines for much of the day [10]. Immediately after the massacre, many miners who had survived the massacre were arrested, making it difficult to find miners to interview. However, as became evident from the UJ team’s interviews and the subsequent Daily Maverick reporting, communication with the miners was not impossible, suggesting the possibility that journalists overstated the difficulties they faced in accessing the alternative account.

The investigative journalism response

Investigative journalism is meant to set itself apart from ordinary reporting in that it uses techniques to dig out information the powerful often want to keep hidden. Such journalism should involve systematic attempts to expose abuses of power, and analyses of how these abuses could have happened in the first place, so that their recurrence can be prevented. In order to play this role, it should also move beyond the humdrum of superficial, day-to-day reporting, and expose the deeper processes at work in society and even anticipate future trends, which inevitably means moving beyond official sources and their understandings of social processes, as these sources may well offer narratives and justifications for their actions that are limited and self-serving (Bromley, 2005: 314–316; Duncan, 2013; Ettema and Glasser, 1998: 7; Konrad Adenhauer Foundation, 2010: 1–5). As a result of its refusal to stick to tried and tested sources, investigative journalism has greater potential to challenge the uncritical reproduction of dominant narratives in the media. In the case of Marikana this would mean moving beyond the police’s argument of self-defence and Lonmin’s argument that the workers’ demands were unreasonable. Furthermore, investigative journalism has more potential to present radical, oppositional accounts about how power structures society, and to whose benefit: accounts that politicise the current social order and make it challengeable by ordinary people. Yet in South Africa, investigative journalism has taken on a particular character, focussing mainly on the conduct or misconduct of political leaders and public servants. This practice of investigative journalism is state-centric, assuming that social change comes about primarily from above and that ‘keeping political leaders straight’ will lead to positive social change. Yet social change can also come from below, and it is this form of social change that journalists have not effectively reported. State-centric investigative journalism is likely to be focussed on effective reforms in the system of oppression and exploitation (that is, achieving capitalism with a human face), rather than presenting fundamental challenges to the system by affirming those social forces that lie outside the state, and that are most likely to exert pressure from below.

Investigative journalist Stefaans Brümmer distinguished between ‘forensic reporting’ and ‘social reporting’, where the former uses investigative techniques to uncover abuses of power, especially by the elite, while the latter lends itself to investigating the causes and effects of social currents [11]. Forensic reporting would involve probing the conduct and affairs of public officials, and uncovering cases of abuse of power; this constitutes the stuff, by and large, of investigative reporting. Social reporting, on the other hand, would include investigations that uncover deeper social trends, such as those related trends in relation to employment and unemployment.

In spite of the fact that this distinction is questionable, it has informed the Mail and Guardian’s reporting, with major ‘forensic’ investigative stories being undertaken by ‘The amaBhungane Centre for Investigative Journalism’, a not-for-profit investigative journalism unit separate from the Mail and Guardian newsroom that provides the paper with investigative stories. However, ‘social’ journalism is often assigned to the much smaller, less well-resourced, Eugene Saldahna Memorial Fund’s fellow in social justice reporting (this is a fellow located in the paper’s newsroom, it externally funded by the Memorial Fund). The paper also offers daily online stories through the Mail and Guardian Online, and this service reported on workers’ allegations that many were killed while fleeing, and some were crushed by nyalas (a South African-made armoured personnel carrier) (Fazel and Sosibo, 2013). This was the first report to put allegations into the public domain suggesting that the police actions were not entirely in self-defence. According to Sosibo, other journalists who reported on the massacre were also told about these allegations by the workers, yet they found these accounts too incredible to report on [12]. This allegation implied the need for a shift from a ‘social’ to a forensic approach, where the allegations were probed and tested for their veracity; however, the paper did not probe this allegation further in its published reporting. According to the Mail and Guardian editor, Nic Dawes, their journalists did try to access hospitals to confirm accounts of the crush injuries, or additional fatalities, but did not manage to find evidence to corroborate the UJ account [13].

Instead, the Daily Maverick online news service took up the challenge of forensic reporting, and analysed the second ‘kill site’ where the bodies were found after having reported on Alexander’s initial allegations. Journalist Greg Marinovich also confirmed in interviews with miners as well as with a confidential source in the police, that the police had, in fact, killed miners at this ‘kill site’. One eyewitness interviewed by Marinovich stated that each time a miner got up and attempted to surrender to the police at the second site, they were shot by the police (Marinovich, 2012b). According to Dawes (2013, personal communication), the Mail and Guardian had invested time and energy in reporting on areas of the country where social pressures were building up and where crucial stories were not being told, including in the platinum belt. When the massacre took place, they were confident that they were ahead of the story as they had invested resources into understanding the events and actors that preceded the massacre. However, the Mail and Guardian failed to tell the story of the second site. Because they could not find independent corroborating evidence of the site, they could not pronounce categorically on the story’s veracity. As a result, they decided to publish the Marinovich piece. Dawes noted that this failure had led to a great deal of self-examination on their part, and while their own efforts, coupled with their previous reporting ‘…did not support a narrative of a metropolitan newspaper able to listen only to the voice of power’ [14], they have also had to confront some of their own reporting weaknesses. According to Dawes:

We also asked whether we have lost some of the skills involved in covering violence that people like Greg Marinovich have built up over years of reporting on conflict? If we have (and the answer is a definitive yes) then how are we going to deal with the increasing importance of protest, sometimes violent, and of police action to quell it (often violent)? President Jacob Zuma made it clear in his State of the Nation address that a crackdown is in the offing, we will need the skills in our newsrooms that we don’t currently have to tell the story well (and to stay safe while doing so) [15].

These frank admissions strongly suggested that while the paper had built up considerable experience in ‘social reporting’, they experienced extreme difficulty in shifting into ‘forensic’ mode when the situation required it. What Marikana underlined was the necessity of being able to merge both reporting modes. Without breaking down the artificial barriers between these two modes, journalists will be unable to rise to the challenge of reporting on any official crackdown on dissent that may be in the offing.

Other more recent events also suggest that the necessary lessons have not been learned from the editorial failure on Marikana. South African journalists largely failed to report on cell phone footage taken by a Tactical Response Team[16] member—showing some of the police actions closer to the second kill site, and capturing a policeman boasting about who he had shot—until it was picked up by Britain’s Channel 4. Another documentary by Channel 4 on police violence in South Africa, alleged that the police responsible for Tatane’s killing could be identified from the screen grabs of footage of his killing shot by a community member (the court maintained they could not be identified as their faces were obscured by helmets) (Gilmore, 2013). Yet there has been little evidence of these allegations having been followed up and tested by South African journalists.

Marikana in the context of the sociology of news

The press reportage on Marikana can be theorised in the context of the sociology of news production, which analyses the factors that shape news production in commercial news organisations. Of central importance here is how journalistic work is constrained by organisational and occupational demands (Schudson, 2002: 249–269). The sociology of news production shows how—rather than being a random response to accidental events—news production is a phenomenon following relatively predictable patterns, and hence is systematic (Schudson, 1989: 263–282). This theory also shows how, far from being free agents, journalists are subjected to multiple and competing internal and external pressures that determine editorial content (McNair, 2001: 62). Here, determination is used in the broad sense of setting limits and exerting pressures  on content, and not in the narrow sense of prefiguring content (Williams, 1973: 3–16).

The decline in labour or industrial reporting has been a worldwide phenomenon since the Thatcherite attack on unions weakened the union movement (McNair, 2001: 62). Labour reporting, in the broader sense of reporting on working class life, has been lost in favour of a form of reporting that episodically reports on major events like strikes. The labour beat has also experienced a decline in status in newsrooms (Talking Biz News, 2011). As a result, specialist bodies of knowledge have been lost from newsrooms. Labour columnist Terry Bell recalls that when he began writing his column ‘Inside Labour’ for The Star Business Report seventeen years ago, almost every publication had a dedicated labour beat. Since then labour issues have been assigned to generalist journalists, who are often young and inexperienced. Had there been more specialist reporters, it would have become more publicly apparent that problems were brewing in the platinum sector for some time, and such reporters would probably have been sent to investigate matters and keep a watching brief on subsequent developments [17]. Yet on the other hand, business journalism has experienced explosive growth in South Africa, leading to the establishment of a number of dedicated titles, and it has also become a coveted beat for journalists (Brand, 2008). Also evident in writings on economic journalism is a conflation of the term ‘economics journalism’ with ‘business journalism’. This implies that economics journalism should be inherently supportive of free market economics, and not open itself up to critiques of this form of economics: a form of socialisation that most likely carries over into newsroom practices (Brand, 2008; Rumney, 2008: 3). Certainly, there were signs in the Marikana reporting that the business case of workers’ demands being unreasonable, and dangerous to the economy, underpinned much of the early economics reporting.

One of the consequences of news commodification from the nineteenth century onwards was that proprietors sought to delink journalism from political parties to create more general audiences for news media. Consequently, they promoted the concept of ‘objectivity’ to rise above the partisan practices of much of the press at the time. Brian McNair has identified ‘objectivity’ as a cornerstone of contemporary journalists’ professional ideology, and has further identified three characteristics of objective journalism: the separation of fact from opinion, a balanced account of a debate and the validation of journalistic statements by reference to authoritative others (McNair, 2001: 68).

While there has been some debate about the concept of objectivity in South African journalism (Bell, 2003), it enjoys wide acceptance as part of journalists’ professional practice, and as a result so too do the strategic rituals attached to this particular concept. The facts that are presented are thus presented as being value-free, rather than containing, in the words of McNair, ‘[structured] bias in favour of the powerful’ (McNair, 2001: 75). In the early reporting on Marikana, many journalists clearly assumed that a limited number of sources were needed for the story to be balanced, and that the workers’ voice was ‘covered’ through interviews with union officials, when clearly it was not. The problem with ‘balance’ is that it can lead to very superficial reporting, where a story is considered to be complete when the two ‘sides’ in a particular story have been canvassed, and both have been given roughly equivalent attention (McNair, 2001: 69). In fact there may be a right and a wrong side to a particular story and, as a result, both sides may not deserve equivalent attention. In such cases, not only is neutrality an inappropriate journalistic response, it is unjust for journalists who may consider it their professional duty to speak truth to power. In the case of Marikana, journalists’ emphasis on balance mitigated against them undertaking a proper investigation, and arriving at appropriate responses.

Marikana also provides a case study of a broader problem in commercial media organisations that tend to take official sources of information much more seriously than the voices of workers or the unemployed. Much of this has to do with the routine practices of news organisations. In fast-paced newsrooms, where journalists are required to meet more and more deadlines, it is tempting to rely on sources of information that are more readily obtainable and have been validated by other media, while avoiding sources that are less ‘trusted’ and require more validation. Known in its more extreme forms as ‘pack journalism’, these tendencies can give journalism a sameness that reduces diversity of voices.

The most easily validated sources are likely to be organisations with the resources to maintain a constant flow of information to the media, such as government agencies, big business and ‘think tanks’. ‘Resource-poor organisations’ (Goldenberg cited in Schudson, 2002: 257) or individuals representing working class or unemployed interests are likely to be less well-resourced and lack the capacity to communicate proactively, which can lead to them dropping under the journalist’s radar. These practices often lead to journalists prioritising dominant groups in society, allowing themselves to become mouthpieces of the rich and powerful, reproducing the official versions of events, and silencing the voices of the workers as rational, thinking beings with their own stories to tell. Apart from these organisational factors, McNair has also pointed out that journalists may also bring cultural assumptions to bear on their selection of sources, assuming that sources whose cultural backgrounds he or she can relate to more easily are more trustworthy (McNair, 2001: 76).

Given the largely middle-class nature of journalism as a profession in South Africa, unacknowledged class biases could also have come into play in the reporting on Marikana. The fact that, on Sosibo’s account, many journalists chose to dismiss workers’ accounts of the full horror of the massacre, suggested that they were less trusting of workers’ accounts than police accounts. Also, the fact that workers associated journalists with officialdom, as argued by Ledwaba, speaks volumes about their perceptions of journalism. As Alexander has noted, workers were more forthcoming with academics. Workers had a history of their stories being distorted and misrepresented by the media, because of their pro-business bias: academics, on the other hand, were seen as being more impartial [18].

News values also strongly affect journalistic practices. McNair has identified two news values that often structure news production in commercial news organisations: deviation and a focus on the doings of elites. With respect to the first, newsworthy events are considered to be those events that constitute disruptions of the norm, like natural disasters, strikes or protests. A focus on these events tends to lead to negative news becoming central to news agendas, with scant focus on what gave rise to the disruption or deviance in the first place, as this information is rarely classified as newsworthy. Elites also occupy a considerable amount of news space, even if they are not acting in deviant ways (McNair, 2001: 78–79). Focussing on elites may well be justified in many cases, as they tend to be the power holders in society; their actions stand to affect many more people than the actions of those who do not hold power. But an inordinate focus on elites can lead to important stories being missed at the grassroots, which was clearly the case in relation to the early reporting on Marikana.

Conclusion: Lessons for journalism

The early press coverage of the Marikana massacre provides a case study of how reporting can become system-maintaining, by failing to take workers’ voices seriously. This silencing of workers’ voices contributed to the disempowerment of the one social force capable of disrupting exploitation in one of the most strategic industries of the economy. As a foreign multinational operating in a highly extractive industry, Lonmin is a beneficiary of the government’s neoliberal turn, and the police appeared to have acted as the company’s private security guards in Marikana. They turned what should have been a public order policing exercise into a counter-insurgency action intent on breaking the strike. Much media reporting also portrays police violence as an aberration, rather than an integral coercive component of a highly unequal, unstable, peripheral capitalist economy. The Marikana case showed just how warped South Africa’s post-apartheid minerals-energy complex has become. Yet little press coverage focussed on the mutually reinforcing interests between the police and the ANC government and Lonmin. As a result, the public was deprived of narratives that brought home the true nature of neoliberalism in South Africa, namely that it is an ideology that justifies super-exploitation, enforced by the barrel of a gun if necessary. The growing problem of police violence is unlikely to abate with the kind of reporting that was evident at the time of the massacre. Indeed, it may intensify as the police may feel that, at best, public discourse is indifferent to the inappropriate use of maximum force, and, at worst, public discourse supports their actions. This problem is likely to grow if the Farlam Commission fails to deliver substantive justice.

It must be acknowledged that there were shifts in the reporting. Once it had become apparent that, over time, there was an alternative account to the police account of events, more journalists included the voices of mineworkers. The ‘Faces of Marikana’ series, a City Press and Media24 Investigations report on the victims of the violence at Marikana, gave the events a personal face. Journalists travelled throughout South Africa to interview families of the victims to piece together their personal life stories. The series included not only the miners, but also the police and the security guard who were killed in the violence preceding the massacre (City Press and Media24 Investigations, 2012). A more recent investigative piece by the Mail and Guardian newspaper interviewed families of those killed and documented the effect on their lives (Tolsi and Botes, 2013). These investigative pieces were important in moving reporting beyond portraying the victims as statistics and focussing on the human costs involved.

However, these stories also confined themselves to portraying workers and their families as victims, failing to assign agency to them. Furthermore, once the Farlam Commission began to sit, journalists became largely dependent on the Commission to tell the Marikana story, at the expense of conducting their own independent but parallel investigations. This suggests that the necessary lessons have not been learnt from the editorial failure in reporting on the massacre, namely that official sources cannot and must not be relied on to tell the story. The Commission is, after all, an official source too.

South African journalism has considerable investigative capacity: that is one of its main strengths. But, trends (albeit uneven) have emerged in how this capacity is being put to use. Those stories that focus on the misdoings of the political elites seem to receive a great deal of attention. Granted, these elites should come in for considerable scrutiny, as their (mis)doings set the stage for abuses further down the chain of command. But, security cluster misconduct that occurs at the point of conflict with the working-class and unemployed is subject to less probing.

This trend reproduces and reinforces broader media and social inequalities. Two decades into democracy, the South African media still constitute an elite public sphere: the number of business publications and business journalists, relative to the number of labour voices attests to this. The unemployed have little voice in the media, except as social problems (such as violent protestors) or as victims. Women and young people continue to be marginalised. This means that media discourses come to the South African public inherently unbalanced. Nothing short of a commitment to social justice, and to affirming society’s most marginalised voices, will correct this imbalance. This should not mean automatically portraying protestors as saints and the state as sinners. But it should mean going out of one’s way to tell stories that those in positions of authority would prefer to remain buried. It also means recognising that those who are not in position of authority are more likely to be bearers of these stories, and seeking them out. It means looking beyond the official spin about protestors as being inherently criminal and therefore untrustworthy.

Why do journalists in a ‘free press’ end up being aligned, often unwittingly, with authority? Sociology of news perspectives hold limited explanatory value. They ascribe particular journalistic choices to organisational and institutional dynamics only, to the exclusion of broader power-dynamics outside newsrooms. As mentioned earlier, the press consider themselves to be news agenda-setters and thought-leaders, but the economic realities in which they operate circumscribe this monitorial role. Furthermore, professional autonomy is less likely in a situation where journalists as a group are weak and unorganised, which is largely the case in South Africa. All these factors limit spaces for forms of journalism that expose how power actually works in society; journalism that reveals systematically how, in spite of two decades of ANC rule, power still largely resides with the very social groups that were advantaged under apartheid. ANC rule has advantaged capital much more than labour, as it has given a democratic face to profit extraction by local and global companies such as Lonmin. Media commodification has also led to much political reporting and commentary being sucked to the political centre, favouring middle-class viewpoints as unacknowledged normative frameworks. This largely elite public sphere in South Africa, typically portrays neoliberal logic often being portrayed as the ‘only game in town’ ideologically, with political alternatives receiving short shrift. Neoliberalism is premised on growing inequality, and in order for it to survive it needs to prevent people from imagining political alternatives. The nature of South Africa’s media system has made the domestication of neoliberalism easier, but has not necessarily led to it penetrating the common-sense understanding of the majority of South Africans.

Political economy concerns itself broadly with the relationship between the structures of control in capitalist society and the production of wealth needed to reproduce that society, including the power relations in the production and consumption of media (Mosco, 2009). There is no evidence to suggest that South Africa’s press owners brought pressure to bear directly on their newsrooms to prevent workers’ voices from coming to the fore in the early press reporting of Marikana, as the mining houses do not own significant stakes in the press. Rather, what was in operation was a set of journalistic rituals that have become all too familiar in commodified newsrooms, and ones that do little to give voice, much less agency, to the most powerless in society. In South Africa, unfortunately, these rituals will do little to confront the most serious national challenge of the post-apartheid period, namely to prevent the descent, once again, into a police state.


[1]        The following is a timeline of events surrounding the Marikana workers’ strike in 2012:

·         10 August: rock-drill operators at Lonmin’s Marikana plant initiated an unprotected strike

·         11 August: a confrontation between workers and the National Union of Mineworkers left two workers injured

·         11–15 August: the strike continues, and ten people are killed in the ensuing violence, including miners, police officers and security guards

·         16 August: police open fire on striking mineworkers, killing 34 and injuring 78 others

·         20 August: most Lonmin workers remain on strike

·         19 September: Lonmin and the workers finally resolve the strike.

[2]        A protected strike is a strike that has gone through conciliation procedures set down by the Act, and if these fail, then workers are protected from dismissal if they choose to strike, while an unprotected strike is a strike that takes place outside these procedures, and which therefore open workers up to dismissal.

[3]        The articles were drawn from a representative sample of South African newspapers, which were the following: Business Day, The Star (including Business Report), The New Age, Citizen, Mail and Guardian, City Press, The Sunday Times and Times, Sowetan, Beeld, Die Burger, Sunday Independent, Financial Mail.

[4]        According to Stuart Hall, primary definers are those sources that initially defined the events and the issues related to the story, and that were instrumental in shaping public perception of the events (Hall et al., 1978)

[5]        Julius Malema and Floyd Shivambu were leaders of the ANC youth wing, the ANC Youth League, but were dispelled after repeatedly criticising the ANC leadership. They have since formed another political party, the Economic Freedom Fighters, which appears to enjoy widespread support in the post-massacre Marikana settlement.

[6]        According to George Gerbner (1972) and Gaye Tuchman (1978), ‘symbolic annihilation’ is the absence of representation, underrepresentation or distortion of particular social groups, with the intention of maintaining social inequality.

[7]        The Benchmarks Foundation is a non-governmental organisation with its roots in the ecumenical movement, and devoted to mining and environmental justice issues.

[8]        Correspondence between Lucas Ledwaba and Ferial Haffajee, 5 March 2013.

[9]        Correspondence between Lucas Ledwaba and Ferial Haffajee, 5 March 2013.

[10]    Ferial Haffajee’s response to author’s presentation on the press coverage of the Marikana at University of Johannesburg, 6 March 2013.

[11]    Personal correspondence with Stefaans Brümmer, 16 January 2013.

[12]    Kwanele Sosibo speaking at Sparks seminar on ‘Reporting South Africa’s policing crisis: Marikana one year later’, Institute for the Advancement for Journalism, 12 August 2013.

[13]    Correspondence with Nic Dawes, 18 February 2013.

[14]    Correspondence with Nic Dawes, 18 February 2013.

[15]    Correspondence with Nic Dawes, 18 February 2013.

[16]    One of the specialised paramilitary policing units of the SAPS deployed to police the Marikana massacre.

[17]    Correspondence with Terry Bell, 5 February 2013.

[18]    Peter Alexander speaking at Sparks seminar on ‘Reporting South Africa’s policing crisis: Marikana one year later’, Institute for the Advancement for Journalism, 12 August 2013.

Author Bio

Jane Duncan is the Highway Africa Chair in Media and the Information Society, School of Journalism and Media Studies, Rhodes University.


Alexander P, Lekgowa T, Mmope B, Sinwell L and Xezwi B (2012) Marikana: A view from the mountain and a case to answer. Johannesburg, South Africa: Jacana Media.

Ashman S and Fine B (2013) The meaning of Marikana. Global Labour Column, March. Available at:

Bell G (2003) Journalists are not objective – Mazwai. Independent Online, 3 June. Available at:

Berg B (2007) An introduction to content analysis. In: Berg B (Ed) Qualitative research methods for the social sciences. Boston, MA: Allyn and Bacon, pp. 238–267.

Brand R (2008) Teaching economics journalism: Why? What? Who for? Paper presented at Journalism Education and Training: The Challenges, Stellenbosch, South Africa, 16–18 October. Available at:

Bromley M (2005) Subterfuge as public service: Investigative journalism as idealised journalism. In Allen S (Ed) Journalism: Critical issues. Maidenhead, England: Open University Press, pp. 313–327.

Capel J (2012) Platinum houses fall short of lofty ideals. The Star Business Report, 14 August, p. 14.

Chauke A and Strydom TJ (2012) Mine war: SA bleeds. The Times, 16 August, p. 1.

City Press and Media24 Investigations (2012) Faces of Marikana. Available at: (accessed 1 February 2013).

Clarke J (2012) Strikes are becoming more violent. Citizen – Money Web Business, 18 August, p. 14.

Crime reporter and SAPA (2012) 18 shot in “battle of Marikana”. Citizen, 17 August, p. 3.

Cwele S (2013) Input for National Treasury Budget Vote (State Security) by Minister of State Security, Dr Siyabonga Cwele. MP National Assembly, Parliament, Cape Town, South Africa, 14 May. Available at:

Dommisse J (2012) Cops ignore “don’t shoot” directive. City Press, 19 August, p. 6.

Duncan J (2013) The stuff of investigative journalism, South African Civil Society Information Service, 14 January. Available at:

Ettema JS and Glasser TL (1998) Custodians of conscience: Investigative journalism and public virtue. New York: Columbia University Press.

Faku D (2012a) Lonmin violence spirals. The Star Business Report, 14 August, p. 13.

Faku D (2012b) Death, illness and debt rock Lonmin. The Star Business Report, 17 August, p. 17.

Fazel A and Sosibo K (2013) Deep read: In the eye of the Lonmin storm. Mail and Guardian, 20 August. Available at:

Fogel B (2013) The selling of a massacre: Media complicity in Marikana repression. 1 February. In: Benjamin Fogel (blog). Available at: (accessed 11 February 2013).

Gentle L (2012) The strike wave and new workers’ organisations: Breaking out of old compromises. South African Civil Society Information Service,. 12 November. Available at:

Gerbner G. (1972). Violence in television drama: Trends and symbolic functions. In GA Comstock GA, and E Rubinstein E (Eds) Television and social behavior, Vol. 1, Content and control. Washington, DC: U.S. Government Printing Office, pp. 28–187.

Gilmore I (2013) Police brutality, and the mass amnesia that threatens to obscure it. Daily Maverick, 22 July. Available at:

Hall S, Clark JN, Critcher C, Jefferson T and Roberts B (1978) Policing the crisis: Mugging, the state, and law and order. London: Macmillan.

Hansen A, Cottle S, Negrine R and Newbold C (1998) Mass communication research methods. London: Macmillan.

Harvey D (2005) A brief history of neoliberalism. Oxford, UK: Oxford University Press.

Hlongwane S (2013) Marikana Commission: NUM in a deep hole over the fight that started it all. Daily Maverick, 1 February. Available at:

Konrad Adenhauer Foundation (2010) What is investigative journalism? Investigative Journalism Manual, pp. 1:1–1:20 Available at:

Laganparsad M and Sutherland K (2012) An eyewitness account. Sunday Times, 19 August, p. 5.

Ledwaba L (2012) The guns of Marikana. City Press, 19 August, p. 4.

Maoto M and Marrion N (2012) Police boss bids to end mine violence as nine die in clashes. Business Day, 14 August.

Marinovich G (2012a) The murder fields of Marikana, the cold murder fields of Marikana. Daily Maverick, 8 September. Available at:

Marinovich G (2012b) Friendship forged in the depths of the earth and strengthened by the horror of Marikana. Daily Maverick, 26 September. Available at:

McKenzie R and Pons-Vignon N (2012) Volatile capital flows and a route to financial crisis in South Africa. AUGUR Working Paper Series 2. Available at:

McNair B (2001) The sociology of journalism. London: Arnold.

Mnyandu E (2012) Leadership is missing in Lonmin’s Marikana crisis. The Star Business Report, 17 August, p. 18.

Moodie G (2012a) Where were the miners’ voices at Marikana?, 15 October. Available at:

Moodie G (2012b) Long term Marikana story dies in ‘war-weary society’., 11 February. Available at:

Mosco V (2009) The political economy of communication. New York: SAGE.

Rees M, Clarke J and Swart D (2012) Lonmin chaos continues as death toll mounts. Citizen – Money Web Business, 15 August, p. 17.

Reuters and Bloomberg (2012) Labour unrest, global growth fears trim JSE’s highs. The Star Business Report, 16 August, p.10.

Rumney R (2008) Editorial opinions: What business journalists need to know. Centre for Economics Journalism in South Africa, 18 September. Available at:

Sacks J (2012) Marikana prequel: NUM and the murders that started it all. Daily Maverick, 12 October. Available at:

Schudson M (2002) The news media as political institutions. Annual Review of Political Science 5: 249–269.

Seccomb A (2012) Two mine guards killed as “inter-union rivalry” flares. Business Day, 13 August, p. 1.

Shoba S and Mahlangu I (2012) Muti “protected” miners. Sunday Times, 19 August.

Sosibo K (2012a) A tinderbox of discontent. Mail and Guardian, 17 August. Available at:

Sosibo K (2012b) Marikana Commission delves into gruesome death of security guards. Mail and Guardian, 6 November. Available at:

Sosibo K (2012c) Farlam Commission: Police tied hands of Marikana wounded. Mail and Guardian, 14 November. Available at:

Sosibo K (2012d) Marikana: Truth, memory and reportage. Mail and Guardian, 21 December. Available at:

Sosibo K (2013) Marikana: NUM witness contradicts himself at inquiry. Mail and Guardian, 30 January. Available at:

South African Press Association (SAPA) (2012a) Nine people killed in mine violence. Sowetan, 14 August, p. 15.

South African Press Association (SAPA) (2012b) Marikana men queued for muti. News24, 8 November. Available at:

South Africa Press Association (SAPA) (2013) Cops “hardened” towards Marikana miners. Independent Online. 25 January. Available at:

Speckman A, Faku D and Mdluli A, (2012) Lonmin halts output as another body is found. The Star Business Report, 15 August, p. 17.

Talking Biz News (2011) The demise of labour reporting. Talking Biz News, 14 January. Available at:

Terreblanche S (2002) A history of inequality in South Africa: 1652–2002. Scottsville, South Africa: University of KwZulu-Natal Press.

Tolsi N (2013) Miners killed like ‘possessed vermin’ says lawyer. Mail and Guardian, 22 October. Available at:

Tolsi N and Botes P (2013) Marikana: One year after the massacre. Mail and Guardian, 16 August. Available at:

Tuchman G. (1978). Introduction: The symbolic annihilation of women by the mass media. In: Tuchman G, Daniels AK and Benét J (Eds) Hearth and home: Images of women in the mass media. New York: Oxford University Press, pp. 3–38.

Vanek M (2012) Platinum belt panic. Citizen – Money Web Business, 15 August, p. 17.

Weekend Argus Reporter (2012) Rabbit blamed for Lonmin deaths. International Online Business Report, 25 August. Available at

Williams R (1973) Base and superstructure in Marxist cultural theory. New Left Review 82: 3–16.