Xenophobic Violence and Struggle Discourse in South Africa

Xenophobia as internalised racism

Some authors argue that xenophobia is a form of internalised racism among black South Africans that has continued from apartheid. Tafira (2011: 114) claims that what is typically presented as ‘xenophobia’ should actually be considered as a form of black-on-black racism, specifically ‘New Racism…practised by people of the same population group, which has characterised post-apartheid South African black social relations’. He argues that ‘people of the same skin colour, in this case black African immigrants and black South Africans, … have over the years been transformed into racialised subjects and … they have come to perceive each other in the light of their racial subjectivities’ (115). For Gqola, too, ‘The specific racialisation of the attacks matters … No one is attacking wealthy German, British or French foreigners in Camps Bay … This is unthinkable. This racialisation makes it safe to victimise Black people from the African continent in our everyday exchanges’ (2008: 213).

Although Tafira (2011) specifically differentiates his ‘new racism’ thesis from characterisations of the May 2008 anti-immigrant attacks as ‘Negrophobia’ (cf Mngxitama, 2008), these accounts all still share an emphasis on the centrality of race and racism in the production of anti-immigrant violence. Tafira uses the concept of ‘new racism’, which emphasises ‘culture’ rather than ‘biology’ in justifications for racial discrimination, to account for how racism can be directed to one group of black people by another. However, Tafira’s new racism thesis still entangles race and nationality insofar as it relies on the citizen-immigrant distinction as the basis for who is victim and who is perpetrator of this racism. Moreover, the paper’s ethnographic data – a collection of names by which South Africans in Alexandra call African ‘foreigners’ – arguably does not substantiate the new racism thesis. The new racism/Negrophobia accounts also do not engage directly with cases of xenophobic violence targeting non-Africans, e.g. Asian shopkeepers (Crush and Ramachandran, 2014a). Indeed, as we shall see below, Monson (2015) argues that race or ethnicity is actually often not the most salient social distinction in the politics of the places where anti-immigrant attacks take place, which rather implicate other kinds of political identities, such as a distinction between the politically committed and the politically indifferent.

Xenophobia is a result of poverty and economic crisis

Some authors explain xenophobia as a consequence of poverty, class formation, inequality and/or neoliberalism. For Amisi et al. (2011: 59), the roots of xenophobia are to be found in structural processes of ‘uneven and combined development’:

a glutted labour market, housing shortages, township retail competition, highly gendered cultural differences, and apparently intractable regional geopolitical tensions. These root-cause pressures continue—as will xenophobia—because short of a national political shift in power and interests, they are extremely difficult to resolve. (59)

By contrast, Von Holdt and colleagues (2011: 90) point to poverty as an explanation for xenophobia via the collapse of social meanings that poverty entails:

where poverty and unemployment are endemic people’s inclinations to subscribe to some lofty societal ideals dissipate. These inclinations dissipate because, abandoned and feeling there is nothing of value for them, people devise ways aimed at giving themselves a sense of worth.

In Amusan and Mchunu’s (2017) formulation, it is inequality specifically which leads to xenophobia: ‘In an environment where some segments of the population enjoy a disproportionate amount of wealth, the inequality within such an environment builds up resentment and eventually, someone has to pay the price for such resentment’ (2017: 4; see also Mamabolo, 2015). Mngxitama also points to racialised inequality in his explanation for the May 2008 violence; he suggests that xenophobic violence is ‘not new’, insofar as it is part of a broader pattern of township violence, which itself is ‘a direct consequence of the wealth of a few’ (2008: 198).

While Fauvelle-Aymar and Segatti (2012, cited in Monson, 2015) did find that the xenophobic violence of May 2008 was correlated with the prevalence of informal housing, a number of authors have pointed out that poverty cannot be more than a partial explanation for anti-foreigner violence, since it does not account for a number of important observations. First, anti-foreigner attacks have not happened in many wards and communities which were at least as poor, and had at least as many immigrants, as the areas where it did happen in May 2008 and at other times (Kirshner, 2012; Polzer, 2010; Von Holdt et al., 2011). Second, economic crisis does not explain why foreigners especially should pay the price for resentment about inequality (Neocosmos, 2008). Neocosmos asks why the rich, or whites, are not targeted instead. Indeed, ‘Poverty can be and has historically been the foundation for the whole range of political ideologies, from communism to fascism and anything in between … [so] how do we logically move from this to stereotyping the foreign other?’ (2008: 586–587). Third, economic crisis does not explain how the ‘foreigner-citizen’ distinction gained its current significance in the first place (Neocosmos, 2006). Accounts that link poverty and anti-foreigner violence would also have to go much further to explain how evicting ‘foreigners’ helps poor South Africans to recover a sense of worth, and why this could not just as well be recovered by other non-violent or non-xenophobic forms of political action. Certainly, anti-foreigner attacks have often been justified in terms of the supposed economic threat migrants pose to the prosperity of South Africans, but this is very different to accepting that this threat is the actual explanation for xenophobia (Crush and Ramachandran, 2014b).

Crush and Ramachandran argue that such economic accounts not only have weak explanatory power but are also politically dangerous. First, because they downplay ‘xenophobia’ as an effect of other, more serious root problems, they offer no real challenge to the pervasive ‘xenophobia denialism’ of the South African government: ‘[xenophobia] minimalists essentially argue that although xenophobia might exist, it is an epiphenomenon that does not get at the root causes of violence’ (2014a: 1–2). Although academics emphasizing the centrality of economic crisis might not be xenophobia denialists themselves, their accounts amount to a kind of ‘xenophobia minimalism’ which Crush and Ramachandran place on the same continuum. Refusals to countenance that South Africa has a xenophobia problem crop up repeatedly in accounts which ‘explain’ attacks on foreign communities either as brought about by things that foreign communities themselves do, or as ‘just crime’.

Second, Crush and Ramachandran are concerned that when the roots of anti-foreigner violence are located in economic competition and scarcity,

then the frames of reference are automatically loaded against [immigrants]. Seen in such terms, resentment and antipathy towards migrants and other “outsiders” become inevitable, inescapable aspects of the social landscape, justifying stringent controls over immigration, and exclusion … of migrants. … [T]his distinction further invigorates the underlying rationale for xenophobia, the very idea that the presence of migrants and refugees poses a perpetual threat to the legitimate insiders (2014b: 13).

Having offered this critique of the poverty/economic crisis account, however, Crush and Ramachandran (2014b) do not provide any alternative explanation of where xenophobia comes from – historical, structural, or otherwise. They simply insist, using attitude survey data, that we must accept as a starting point that South Africans – of all races and classes – are highly xenophobic, a position they call ‘xenophobia realism’. They highlight three survey findings in particular which indicate strongly xenophobic attitudes:

(a) the nature and strength of myths about migrants and migration; (b) the level of public endorsement of coercive state measures to keep migrants out of the country and to remove those who are present; and (c) the degree of willingness to resort to coercion and violence against migrants (18).

Crush and Ramachandran’s point is that the xenophobia of South Africans, not economic crisis, is the reason for anti-foreigner attacks – hence their refusal to look elsewhere for ‘causes’.

Xenophobia as a byproduct of new South African citizenship/nationalism

Neocosmos (2006) offers a historical explanation of why xenophobia exists at all in South Africa, and how it emerged in the transition from apartheid. He emphasises the changing meanings of citizenship in the production of xenophobia. Neocosmos argues that the current strong distinction between black people who are South African citizens and black people who are non-citizens or ‘foreign nationals’ is something that only emerged in the transition from apartheid. This distinction was not made strongly during apartheid either by the state or by the anti-apartheid movement, because the state was attempting to ‘denationalise’ all black people anyway by making them citizens of separate homelands, and regulating their presence in ‘white South Africa’ according to how it could make use of their labour. In this way, black people in South Africa were rendered both foreign and migrant (Pillay, 2013), and treated much the same as foreign migrants from further afield. Indeed, the Transkei, Ciskei, Venda and Bophuthatswana homelands were actually modelled on the neighbouring ethnic nation-states of Lesotho, Swaziland and Botswana. This ‘mode of domination’ meant that

bonds of solidarity were developed between all Africans in the region and beyond, so that the struggle against apartheid was very much conceived … as a fight of all Africans and their allies against the apartheid state. The concept of “nation” thus developed tended to be inclusive rather than exclusive of Africans from the region (Neocosmos, 2006: 30–31).

While there were other forms of ‘black-on-black’ violence during especially the latter years of apartheid, these occurred around political, ethnic or migrant identities, such as the African National Congress-Inkatha Freedom Party conflict in Natal/KwaZulu and the Transvaal (Minnaar, 1992); they were not framed in terms of a distinction between who was and was not a South African citizen. So xenophobia was not an inevitable outcome for post-apartheid South Africa, but, as the new state chose to treat its population as ‘passive … subjects of state or white largesse’ (Neocosmos, 2006: 77), a distinction then needed to be made between those living in South Africa who qualified for state assistance and those who did not. The dividing line for this distinction was new South African citizenship:

Under apartheid all rural migrants whether emanating from South African territory or not, were interpellated as foreign through the medium of tribal identification. Post-apartheid, only those emanating from beyond South Africa’s borders are interpellated as foreign, as the Bantustans are simply struck off the map. It is no longer ethnic identity but national (and increasingly black African) identity which enables access to resources. (19)

Thus, a new set of reconfigured borders delineating foreigners from citizens replaced the old ones; and simultaneously a new category of denationalised outsiders was created amongst those living within South Africa’s borders. Thus, for Neocosmos, the transition was ‘a transition between two different forms of xenophobia, simultaneously with continuity between state practices’ (2006: vi). Neocosmos’s account thus fills an important gap in the new racism, economic crisis and ‘xenophobia realism’ accounts by pointing to the historical relationship between citizenship and the ability to make claims on the state. However, as we shall see below, the connections between xenophobia and citizenship have been nuanced by findings that xenophobia is often justified by South Africans who see themselves as also excluded from the rights of citizenship, rather than ‘an exclusionary claim by those who already belong’ (Monson, 2015: 149; Solomon, 2019).

Anti-foreigner violence is a consequence of a weak state and political expediency

A key insight made repeatedly by scholars at the African Centre for Migration and Society (ACMS) is the role that local political actors – either ward councillors or informal, sometimes self-elected leaders – often play in inciting xenophobic violence for their own political and economic gain (Misago et al., 2010). Appropriating shops, goods and houses and being seen to be effective at solving community problems are some of these gains. The ACMS work has been driven by a conviction that ‘violence … is a form of conflict with its own dynamics’ (Brubaker and Laitin, 1998, cited in Misago, 2009: 13) and which requires explanation. Since xenophobic attitudes do not always or automatically translate into violence, this research has aimed to identify the ‘trigger’ factors by which whole communities are mobilised to attack foreigners. This work also takes us beyond global, deterministic accounts of context ‘causing’ xenophobia and into a local, blow-by-blow account of how xenophobic attacks are actually planned, incited and executed – often at public meetings, sometimes by ward councillors, sometimes preceded by the issuing of ‘eviction notices’, and often coinciding with service-delivery and other kinds of protests (see also Palmary, 2016; von Holdt et al., 2011). Polzer argues that xenophobic attacks happen where the state is absent or compromised and that ‘violence against foreign nationals and ethnic minorities is a symptom of broader challenges of legitimate and accountable local governance, especially in informal settlements’ (2010: 5). This work thus locates xenophobic violence in the ‘micropolitics of township and informal settlement life’ (Misago et al., 2009: 10). In this way, it appears that conditions of poverty and a weak state associated with informal settlements are often a contributing, but insufficient, condition for anti-immigrant violence.

This work moves us beyond determinism because it breaks the automatic link between pre-existing contextual factors and the emergence of violence. However, if we are to see anti-immigrant violence as more than only an ‘elite’ project (Monson, 2015), we need some account of xenophobia that is popularly diffused. Presumably incitements to violence must resonate with what Crush and Ramachandran (2014b) identify as the already-existing and very widespread xenophobic attitudes of South Africans towards ‘foreigners’ if these incitements are to be effective. Consider how King Goodwill Zwelithini’s infamous 2015 ‘moral regeneration’ speech, in which he called on foreigners to ‘take their bags and go back where they came from’, and on the South African government to ‘help us rid our land of lice’ – even while invoking, in the same breath, Southern Africa’s shared history of liberation struggles – was met with loud cheers by the audience (Zwelithini, 2015). Misago et al. (2010: 10) do acknowledge that local leaders have ‘used popular frustration’ to drum up support for violence, but, as has been argued elsewhere (Kerr and Durrheim, 2013), some authors have tended to dismiss these frustrations as being based on unfounded myths about the economic threat that immigrants supposedly pose (e.g. Crush and Ramachandran, 2014b; Misago, 2009). Understandably, many scholars take an anti-xenophobic position and are loath to take these ‘frustrations’ seriously lest they appear to be simply repeating xenophobic beliefs about migrants taking South Africans’ jobs and so on.

Is a social psychology of xenophobia possible? Material conditions are mediated by systems of belief

Henri Tajfel, a social psychologist who worked in the UK between the 1960s and the 1980s, argued for a social psychological account of group conflict that could explain the relationship between people’s material conditions and their collective choices of action: ‘A psychological theory of intergroup relations must provide a two-way link between situations and behavior’ (1981: 129). The way it can do this is with an analysis of the meaning-making processes that intervene between the two. In so arguing, Tajfel was challenging the determinism of many psychological and sociological explanations, which locate(d) the origins of group conflict purely in competition over resources, or in evolution, or in the attitudes of the subconscious mind – as if these psychological or structural factors exert a unidirectional influence on people and compel them to commit certain actions. In social psychology, Tajfel’s own discipline, deterministic explanations for group conflict still proliferate, often taking the form of circular reasoning whereby analysts observe group- or identity-based conflict and then conclude that this must be because of some natural (‘psychological’) propensity for people to identify with groups and to engage in group-based conflict (Reicher and Haslam, 2013). These explanations fail because they have little to say about cases where conflict is absent despite the existence of group identities, or about how the salient identities around which conflict occurs at one historical moment have ceased to exist or be important at other moments (Reicher, 2004). Instead, Tajfel argued that ‘the best way to predict whether a person will harbour hostile attitudes towards a particular group and what will be the content of those attitudes is to find out how he [sic] understands the intergroup situation’ (1981: 130). As we have seen from the review above, however, explanations for xenophobia have not generally shown much interest in how South African citizens who support or commit anti-foreigner violence understand the ‘intergroup situation’ between themselves and immigrant African communities. This may be because academics are wary of taking extremely xenophobic ‘myths’ about migration and immigrants seriously; or because sometimes academics’ own academic/political commitments incline them towards paradigms which offer explanations in advance of an investigation of local meanings (e.g. Marxism). While Tajfel acknowledged that ‘it would be no less than ridiculous to assert that objective rewards (in terms of money, standards of living, consumption of goods and services, etc.) are not the most important determinant of … [many intergroup] conflicts’, he insisted that another part of the picture are the systems of beliefs about groups (which he sometimes referred to as ‘stereotypes’) that mediate between material conditions and the way group relationships actually unfold:

These ideas, attitudes and systems of belief become an inherent part of the intergroup social situation and …, in a variety of conditions, they are able to deflect in one direction or another the course of the relations between the people involved. … [N]ot only the origins of stereotypes but even their contents cannot be dissociated from the prior existence and the special characteristics of a conflict of interests. However, once they are in existence, they become in their own right one of the causal factors which needs to be taken into account in the analysis of intergroup relations (1981: 225).

Tajfel was thus gesturing towards the ‘productive’ relationship between structure and agency: conflict is not reducible simply to economic competition, because ‘stereotypes’ or meanings about the collective self and the collective other intervene, producing group relationships whose direction and meaning take on a life of their own, or become greater than the sum of their parts. ¹ This formulation overcomes determinism because it accounts for the possibility of change by presenting a reciprocal relationship between context and sense-making that develops over time.

More recently Durrheim and Dixon – also social psychologists of intergroup relations – took up Tajfel’s same question, calling on social psychologists to pay attention to ‘the meanings that participants themselves attribute to their encounters with others’ in particular socio-economic contexts (Dixon et al., 2005: 701). Dixon and colleagues were responding primarily to a social psychological research tradition called the ‘contact hypothesis’, in which researchers seek to measure the quality of interracial or intergroup contact in desegregated societies in terms of a set of pre-determined dimensions (e.g. Islam and Hewstone, 2003) – in the process nurturing ‘a strange incuriosity about how participants themselves make sense of their encounters with others within particular sociohistorical circumstances’ (Dixon et al., 2005: 701). As an alternative to the contact hypothesis tradition, Durrheim and Dixon proposed that researchers investigate local meanings and understandings of contact and of the nature of collective relationships from the perspective of the people participating in them. In what terms do people construe the nature of ‘us’ and ‘them’ and, importantly, the relationship between the two? For this purpose, they proposed the concept of ‘working models of contact’. Group members employ ‘a common frame of reference in which contact comes to be understood in terms of consensual meanings and values. We will call such shared frameworks of explanation and evaluation “working models” of contact’ (2005a: 66). Contextual conditions of poverty, anti-black racism, local politics and the emergence of a new kind of post-apartheid nationalism may inform such models of contact between South African ‘citizens’ and ‘foreigners’, but they do not totally determine them. Also, such models draw on ‘myths’ and ‘stereotypes’ about immigration and immigrants, but are not completely reducible to them, and they cannot be dismissed as simply ‘false’. Rather, they render sensible the instantiation or enactment of beliefs about the nature of the intergroup relationship (Durrheim and Dixon, 2005b); and so are ‘not merely interpretive frameworks for “making sense” of social relations; [but] are also performative explanations that service a range of social, rhetorical and ideological ends’ (Durrheim and Dixon, 2005a: 68). In the rest of this paper, we want to analyse how, and with what social and political consequences, such explanatory practices accomplish or challenge xenophobia within particular contexts.

Case 1: Xenophobic violence averted in Khutsong

Kirshner (2012) offers an analysis of how anti-foreigner violence was averted in the township of Khutsong, next to Carletonville on the West Rand, in May 2008. He highlights the role played by a local community-based organization, the MDF, in discouraging xenophobic violence. Kirshner also constructs a history of struggles between community and municipality – in this case, community resistance between 2005 and 2007 to the state’s attempts to incorporate Khutsong into the North-West province from Gauteng. Khutsong had a history of ‘low-level xenophobia’, and there had been a Mozambican community there since the 1950s – initially living in mine worker compounds, but from the mid-1990s more integrated into the township community. Khutsong residents took part in violent ‘anti-demarcation’ protests against changing the municipal boundaries between 2005 and 2007; the mayor and ward councillors were largely absent from these. But when the anti-foreigner violence of May 2008 broke out elsewhere in South Africa, other community leaders actively encouraged Khutsong residents not to follow suit:

Khutsong residents witnessed the disquieting events [of May 2008] on the TV news. As the crisis unfolded, the local community leaders invited members of the community to Khutsong Stadium, as they had done repeatedly from 2005 to 2007. ‘They told us not to be afraid because of what was happening in the townships outside Johannesburg, in Alexandra. They said whatever was happening in Alex would not happen here,’ recalled a Mozambican man who has lived in Khutsong for 9 years. (2012: 1314–1315)

Kirshner suggests a number of reasons for this active discouragement of xenophobia; some strategic, and some principled. Anti-foreigner attacks may have detracted from the demarcation protests; MDF organisers actively downplayed nationality-based divisions, instead emphasizing local provincial boundaries (‘We are Gauteng people’) and the importance of ‘a collective sense of place around which opposition to the government’s plans could build’ (1320). Importantly, Kirshner shows how leaders mobilised against xenophobia by invoking struggle discourse, which was used inclusively rather than divisively:

We said ‘these are our brothers and sisters.’ We told people that blaming foreigners was morally wrong. After all, we owe Africans a debt of gratitude as repayment for their sacrifices during the anti-apartheid struggle and for their contribution to building our mining industry. … Zuma stayed in Mozambique during the liberation struggle, and Chris Hani was in Zimbabwe. We made these examples to show people we have a history with neighbouring countries in the struggle. (Mogale, cited in Kirshner, 2012: 1320)

Kirshner therefore argues specifically against accounts of xenophobia which treat it as an automatic consequence of poverty and deprivation. Instead, his study ‘highlights the importance of examining local social struggles and their intersections with broader political-economic dynamics in explaining the presence [or absence] of xenophobia in specific places’ (2012: 1324).

Case 2: Anti-xenophobic activism in De Doorns

In De Doorns, there were different kinds of anti-xenophobic responses to the eviction of the Zimbabwean community in 2009 (Kerr, 2017). One came from farmers, some of whom strongly condemned this attack and its instigators, and who offered accommodation to displaced Zimbabwean workers afterwards (Kerr and Durrheim, 2013). Another came from Zimbabwean workers themselves, who dismissed the claims that they had been working for less than the minimum wage and instead pointed to South African workers’ own shortcomings for why they were being overlooked by farmers (Kerr, 2017). The dilemma of these kinds of responses, however, is that they may simply repeat the same ideological divides which were taken as a justification for violence in the first place, as those who opposed xenophobia were also seen to be anti-struggle.

Another kind of anti-xenophobic politics, however, was practiced by mainly Zimbabwean members of the migrants’ rights group PASSOP, who were active in De Doorns after the Zimbabweans were evicted in 2009, up until, and during, the farm workers’ strikes that took place around De Doorns in late 2012 and early 2013 (Wilderman, 2014). During the strikes, PASSOP organisers had to tread a very fine line between the moral universe of worker liberation and the moral universe of anti-xenophobia. They had to find ways of challenging the often highly xenophobic, divisive tactics of some South African strike organisers, while simultaneously showing that Zimbabweans, too, were actively participating in workers’ struggles (Kerr, 2017). PASSOP also highlighted the common class identity of all workers as workers, and tried to downplay the national and racial divisions that had featured strongly in earlier anti-immigrant violence (Kerr, 2017). This is another example of how struggle discourse was mobilised in inclusive rather than divisive ways. PASSOP organisers claimed they had been reasonably successful, insofar as there were no attacks on Zimbabwean farm workers during the farm workers’ strikes. ²

Case 3: Abahlali baseMjondolo’s anti-xenophobic struggle in Durban

Abahlali baseMjondolo, the shack dwellers’ movement that was formed in Durban in 2005, is locked in an ongoing struggle with the ANC government in KwaZulu-Natal over housing, land and the freedom to organise independently of the ANC (Abahlali baseMjondolo, 2018; Clarke, 2018). In the course of this struggle, Abahlali have taken a consistently anti-xenophobic position. The movement began when a piece of land set aside for new RDP houses in Durban was sold to a private buyer, and Abahlali’s ongoing grievances concern the corruption that prevents proper allocation of state housing to the poor (Clark, 2018). An enduring theme of Abahlali’s writings is the persecution and victimization of black, poor people in South Africa by the state, irrespective of where they come from:

When we try to unite and to take to the streets to assert that every person is a person, that everyone counts, we are openly beaten by the police. Once again we say that there is no democracy for the poor in this country. It does not matter which country you were born in, or what part of South Africa you come from, or what language you speak. If you are poor and black you are excluded from this democracy with the open use of violence (Abahlali baseMjondolo, 2015).

They point out that, while the anger of poor South Africans is justified, turning on their foreign neighbours is not going to solve the problems of ‘rats, fires and lack of toilets’ that they experience (Abahlali baseMjondolo, cited in Solomon, 2019: 166). In this way, ‘Abahlali directly challenges the exclusive articulation of rights underlying [xenophobic] arguments that South Africa’s poor are the victims of migration’ (Solomon, 2019: 167). Moreover, Abahlali have ‘used their experience of being marginalized as “illegal” squatters to challenge the distinction between citizens and noncitizens in the country’ (164). For instance, they draw attention to similarities between informal settlements, refugee camps, repatriation centres and other ‘dumping grounds’ for unwanted people. In this way, Abahlali have reconfigured the boundaries between who belongs and who does not by emphasizing not formal citizenship – from which they themselves are in practice excluded – but rather by building solidarity among the excluded poor. In this way, Solomon notes that ‘by calling themselves the “people of nowhere” … Abahlali’s anti-xenophobia work challenges the prevailing geography of citizenship’ (167).

This anti-xenophobic position has come at a cost, however:

We have been working to build a politic from below that accepts each person as a person and each comrade as a comrade without regard to where they were born or what language they speak. In this struggle we have faced constant attack from the state, the ruling party and others (Abahlali baseMjondolo, 2015).

During anti-immigrant attacks around South Africa in April 2015, Abahlali described the punishment they received for their anti-xenophobic action in Durban:

On the 8^th of April we supported a march against xenophobia organised by our comrades in the Congolese Solidarity Campaign together with the Somali Association of South Africa and other migrant organisations. There was a permit for the march and yet the police would not allow it to go ahead. They stopped people from leaving their communities to travel to the march. They attacked the march with tear gas, water cannons and rubber bullets. One Congolese man was severely beaten by the police with a plank. One of our members, from the Marikana Land Occupation in Cato Crest, had her leg broken during the assault by the police. We also noted senior police officers accusing Abahlali: ‘What do you have to do with this march? Why are you supporting them?’ … At the march on the 8^th of April there was another march of the so called ‘locals’ who were screaming and saying ‘awahambe’ (‘foreigners must go’). … We were not only assaulted by the police. We were also threatened and assaulted by this group who said to us: ‘Why are you supporting these foreigners’. On that day the police were supporting this group. Despite the violence and intimidation from the police and ‘the locals’ we made it to the City Hall. (Abahlali baseMjondolo, 2015)

In this way, Abahlali pursue a very different kind of struggle than that described earlier in this paper. Rather than claiming that ‘foreigners’ undermine the struggle, Abahlali call for solidarity among the poor on the basis of shared rightlessness.