From Violent Policies to Policies for Violence Prevention: Violence, Power and Mental Health Policy in 20th Century South Africa

Butchard, A., Hamber, B., Terre Blanche, M. & Seedat, M. (1998). From Violent Policies to Policies for Violence Prevention: Violence, Power and Mental Health Policy in 20th Century South Africa. In Foster, D. Freeman, M. & Pillay, Y. (eds.), Mental Health Policy Issues for South Africa, Cape Town: Medical Association of South Africa.

 

In Foster, D., Freeman, M. & Pillay, Y. (eds), Mental Health Policy Issues in South Africa, pp.236-262, Cape Town: Medical Association Multimedia Publications, 1998.

Alex Butchart is Deputy Director of the Institute for Social and Health Sciences at the University of South Africa.

Brandon Hamber is an independent consultant.

Martin Terre Blanche is a Senior Lecturer in the Department of Psychology at the University of South Africa.

Mohamed Seedat is Director of the Institute for Social and Health Sciences.

Deliberate strategies for the use or containment of violence have long played a substantially greater role in the dynamics of South Africa than they have in many other societies. As a result, references to the violent practices and policies of the past are prominent in the popular and political rhetoric of the late 1990s, and contemporary efforts to deal with the problem of violence are shadowed by what went before them. To increase the likelihood of their effectiveness, new policies should therefore be formulated with reference to what these historical strategies (and in particular those developed in the course of the twentieth century) may teach us about the management of violence in the present. The context within which new policy initiatives come about is however not limited to explicitly formulated strategies and programmes, but also includes "infra-policies"(i.e. loosely defined discourses, ideas and practices) that may or may not converge with formally stated plans of action and strategies of response. Accordingly, this chapter's policy proposals are presented against the background both of broad historical tendencies in twentieth century South Africa, and of formal policies and policy proposals concerning violence. In reviewing these policies and infra-policies from a mental health perspective, it is important to note that while the psychological sciences have always occupied a central position in giving intellectual shape to violence in South Africa, this has always been in close inter-relationship with other approaches to the management of violence, such as the police and criminal justice systems, the military, and the public health apparatus.

This chapter has three aims. First, to place violence in South Africa and social responses to it within the interpretive framework of Foucault's theory concerning power and socio-medical knowledge, and to identify the implications that this structuralist approach has for policy formulation. Second, to outline the four major discursive regimes that have conditioned South African practices in relation to violence over the course the twentieth century. Third, to present a set of concrete policy recommendations that appear congruent with the latest discourse on violence and are grounded in a public health practice designed to enable intersectoral collaboration around the prevention of violence in South Africa.

Violence in Twentieth Century South Africa: A Foucaultian Perspective

Violence, like every other issue to come within the attention of the socio-medical sciences, exists not above but within history, and not merely as an object of attention but also as a strategy of perception by which the psychological and the social are made visible and rendered manageable. From the viewpoint of Foucault's theory of power (Foucault, 1977; Gordon, 1980; Kritzman, 1988), the importance of such an observation is twofold. First, it recognises that violence cannot simplistically be seen as a destructive force alone, since counterpointing its damaging effects are the productive consequences of attempts to analyse and understand its meaning and origins, by which the identities and attributes of people, groups and societies are continuously invented, sustained and transformed. Second, it highlights the trans-humanist forces that delimit the boundaries of knowledge and action to create a non-progressive series of historically distinct patterns of human and social response. For, encoded in "discursive regimes", these force fields condition at any point in time what it is possible to see, say, think, and do. In relation to violence, this means that the range of possible policy options in the present is always restricted to the confines of the prevailing repertoire of violence, and therefore that any system of response is in fact a reflection not so much of the will of the people who set it in place, as of the unpredictable play through them of a power that is both "intentional yet non-subjective" (Foucault, 1979, p.94). Precisely because it problematizes the liberal-humanist belief in individual agency and the inevitability of progress, Foucault's theoretical perspective offers a useful counterfoil to this chapter's discussion of past and future policies aimed at self-consciously directing our response to violence.

According to Foucault, a productive relationship exists between power and knowledge at any particular time, with the result that an apparently objective phenomenon such as violence is in fact fabricated in historically contingent ways as an outcome of this relationship. In Discipline and Punish Foucault (1977) explicitly examined the role of violence in relation to power. He argued that over the past 300 years there occurred a shift in the power-knowledge relationship in the West from sovereign power, where social control emanates from a central nexus of authority, to disciplinary power, through which social control is dispersed throughout society and down to the finest grain of individual and sub-individual determinants of behaviour. Under sovereign power, knowledge of the state and respect for its might are inculcated through highly visible displays of the state's capacity to command obedience, the most prominent of these being actual and symbolic shows of force and the capacity to inflict violence on the citizenry (e.g. public executions; military parades; chain gangs; news footage of police or army action). Disciplinary power, by contrast, works by directing attention onto the individual subjects of the state, and through such individualizing and subjectifying technologies as the survey, clinical medicine, the psychological interview, the public opinion poll, and training in the ways of physical and mental health, operates to recruit them into becoming their own overseers, exercising power over, and against themselves. Sovereignty produces both the 'memorable man' as the custodian of power and the oppressed masses who sporadically try to wrest it from him; discipline produces the 'calculable man', this being the modern individual who believes in individual autonomy and who directs his or her agency to caring for and enhancing that agency in ways that conform to the greater good of society.

Although sovereign power in all but its symbolic form has disappeared from most Western countries, the dynamics of South African society and politics in the twentieth century can be interpreted in terms of both sovereign and disciplinary power, local policies and practices in relation to violence having been conditioned by both these modalities of power. Along with the rise of modernity as it was enabled by a shift from the brute force of colonial sovereignty to a more disciplinary mode of colonialism, violence first became a target for socio-medical investigation in the late 1890s, and from then until around 1960 was understood in predominantly psychodynamic and ethnopsychiatric terms. In South Africa (Butchart, 1995), as elsewhere in Africa (McCulloch, 1995; Vaughan, 1991), the "African mind" was thus fabricated as a calculable object of intervention, this regime at once pathologizing violence and sustaining attempts to prevent it through the manipulation of traditional tribal structures and values. In the 1960s, this ethnopsychiatric vision of violence began to fade within a new era of sovereign power for which violence (both as symbol and act) was wielded as a legitimate tactic of political struggle by the apartheid state and its opponents. From being a sign of psychopathology, violence became a signal of political might, the "moral orthopaedic" practices of socio-medical analysis and intervention being overshadowed by the sciences of war and revolution to install a reign of terror. A third important shift in how violence was understood and responded to commenced in the late 1980s. Contemporaneous with the collapse of apartheid and moves towards democracy, so the use of violence for sovereign ends was deligitimized. Now the concern was not the spectacle of violence as a means to political power, but rather the consequences of violence – for the individual, the family, the community and society. Propelled to a heightened visibility by the vigilant attention of the socio-medical sciences and the guardians of human rights, it was now the victim of violence which dominated within a new disciplinary regime preoccupied with documenting and healing the physical, psychological, and social wounds inflicted by the political violence of the past. While in 1996 the response to violence continues to be a predominantly reactive one dominated by the cult of the victim, there can be discerned over the last five years the beginnings of a new and distinct disciplinary recognition of violence. Known as the sociological, demographic or public health approach, this is as much concerned with remediation as with prevention. Accordingly, it constructs violence as the product of a vast web of socio-ecological relationships and risk factors that impinge upon people to increase or decrease their proclivity for violence, and which through adequate identification by way of epidemiological and social research can be manipulated to prevent the problem.

(i) Basic instincts and moral orthopaedics: violence as a psychodynamic phenomenon

The late nineteenth century rise of industrialisation in South Africa brought together large numbers of Africans and Europeans in the confined urban spaces of the mine, the factory, the shop and the home, and it was in response to these problems of proximity that the problem of violence for the first time became an object of concern for the human sciences in South Africa. Prior to this point in the 1890s violence, had, of course, been systematically applied as a tactic of colonial expansion and weapon of African resistance from the 1600s onwards, as exemplified in accounts of life on "the frontier", that boundary zone between European "civilisation" and African "barbarism" in which battles were fought and the spectacle of the gallows played out (see De Kock, 1950; MacCrone, 1957). By the 1890s, however, the new social configuration imposed by industrialisation demanded a radical transformation in the discourse of violence. For the old ways in which it had been used as a weapon of sovereign command through terror were no longer practicable, their efficiency undermined by a great swelling in the size and density of the population living in towns and cities.

What this required was a more efficient arrangement of the capacity to control people, the installation of a power that worked less upon the body, than into and through it by way of analysing the body's inner dynamics and transforming its previously irrelevant urges and desires into malleable objects of a management strategy directed to the manufacture of "docile" bodies. In broad terms, this marked the dawn of modernity and a shift in South Africa to what Foucault termed "disciplinary power". Now, violence for the first time passed over and into the domain of the human sciences, this point of transition from the memorable man to the calculable man somewhat dramatically marked by a new fear in the hearts of Europeans as to the dangerous "instincts" and "passions" that lurked in the dark of the newly invented African mind.

In 1893 a Johannesburg morning newspaper could thus carry the following somewhat dire warning to its readers: "Beware of your houseboy, for under the innocent front may be lurking and lying latent the passions of a panther, or worse!" (cited in Van Onselen, 1982, p. 49). This response to the attempted rape by a black male servant of his white "madam" catalysed the first of a series of "black peril" scares that traversed the Witwatersrand between then and 1912, and in the shape of the "swaart gevaar" and the "rooi gevaar" would resurface at times of tension well into the 1980s. As Van Onselen (1982, pp. 50-53) has shown, the majority of actual incidents of violence around which the 1893 episode crystallized had clustered among economically less stable working and middle-class households, suggesting that as a collective phenomenon the "curse of the black peril" was driven by "periods of stress and acute tension within the political economy … as a whole" (p. 51).

With the lunacy amendment act of 1908, the "basic instincts" underlying the curse of the black peril found their first formal inscription in a policy document that distinguished between "Governor's Pleasure", "Criminal" and ordinary "lunatics" (Transvaal Archives Depot (TAD), CS 863/14966), to delineate all those who fell into the intersection between madness and badness as "dangerous individuals", a category of deviance that in this colonial context was to express itself with particular clarity in relation to violence and the African. Thus, by 1910 the impulsive African as a dangerous individual found further formalization in the Secretary for Native Affairs' suggestion that asylum superintendents endorse the passes of all native lunatics as to the fact of and duration of their detention in the asylum, and ensure that they were repatriated to their own kraals, not only to purify the white community of their threat, but also "so that the patients may have a complete rest … to strengthen their recovery" (TAD, 202/10 LD 1786).

Two years later, this rapidly consolidating psychodynamic discourse on violence found further confirmation in a 1912 editorial in the South African Medical Record (Editorial, 1912, p. 201) which called for "the curse of the black peril" to be dealt with through: "a scientific treatment of this social evil on exactly the same lines as if it were a human disease". Locating the cause of "the black peril" to the removal of Africans from their tribal environment, it observed how:

We have taken enormous hordes of young adult savages or semi-savages, eminently virile in more senses than one, from their own environment, and have placed them in an environment absolutely teeming with every possible stimulus to the sexual impulse at the same time that they are, necessarily, kept celibates … . We have not even tried to put them in the social mosquito-proof house of a reproduction of a native community, but, on the contrary, have freely exposed them to all the stings of a class of human mosquitoes whose interest is to inoculate them with every kind of human vice, and, as regards some forty thousand of them at least, have permitted their employment in duties of all other most calculated to raise the sexual impulse. (Editorial, 1912, p.203)

Consolidated in the 1920s by the burgeoning of academic disciplines (e.g. anthropology, ethnology, psychology, sociology) concerned with human and in particular African thought and behaviour (see Dubow, 1987; Foster, 1990), and by the infamous analyses of the Carnegie Commission into the poor white problem (e.g. Albertyn, 1932), the discourse of violence as a psychodynamic problem had by the 1930s embraced the African in a web of infra-policies aimed at preventing violence – both among Africans, and by Africans towards Europeans. Confirming that violence was somehow related to madness, and madness in turn to urbanisation, Sachs' Black Hamlet (1935), and Laubscher's Sex, Custom and Psychopathology (1935), demonstrated the problem of violence to be especially prevalent in the towns and cities. Hence a system of moral orthopaedics focused on remedying the urban African's lack of super-ego control by recreating carefully selected aspects of tribal culture in the urban native location.

As Laubscher noted: "the commission of sadistic sexual acts on European women by detribalized natives living in towns led me to enquire into the frequency of the occurrences of sexual offenses in the native territories" (Laubscher, 1937, p. 257). Finding such offenses rare among rural Africans, Laubscher concluded that their prevention in urban settings could be achieved through manipulation of the tribal rites relating to African masculinity. Thus, because neglect of the circumcision rite created "a marked instability … in behaviour and attitude to practical things", it should not only be permitted but actively encouraged among urban Africans (p. 134). In contrast to this stabilising and therefore desirable rite was the "racial characteristic of sharing and mutual assistance" (Laubscher, 1937, p. 135). In the rural areas this was an admirable attribute. In the towns, however, it "facilitates his comprehension of communistic ideals … and makes the native prone to the influence of agitators" (p. 196-197). Such altruism was therefore to be met with the antidote of education, carefully calculated to engender a desire for private property within the framework of those racial and tribal traditions consistent with "capitalistic administration" (Laubscher, 1937, p. 197).

Elaboration of this psychodynamic approach to violence would continue well into the 1970s, when building on the psychometric tradition of moral and personality surveillance established by Biesheuvel (e.g. 1953; 1955; 1957), Sherwood (1957), and De Ridder (1961), the last psychometric device explicitly designed to monitor the individual's tendencies to violence was published by the Human Sciences Research Council as the "Zulu TAT" (Erasmus, 1975). However, the disciplinary thrust to monitor violence through such psychological surveillance and the screening out of dangerous individuals (Editorial, 1968) was by the 1960s subordinate to a new sovereignty in which the use of violence by the state and by its opponents eclipsed the contribution of the psychological sciences to its control, all but obliterating the visibility of ordinary individuals as victims and perpetrators of violence.

(ii) The New Sovereignty: State Violence and the Armed Struggle

Foucault's famous description of the 1757 execution of Damiens the regicide (1977, p.3) was a reminder that the mark of a society in the grip of sovereign power is the strategic deployment – both by the state and by its opponents – of violence as a mechanism by which the formal centres of control and resistance display themselves to the public eye, asserting and confirming their might through their capacity to create spectacles of death and destruction. It is therefore to 1960 that the first great rupture in twentieth century South Africa's discourse on violence can be located. For it was then that intensification of the armed struggle as a strategy of African resistance against the oppressive policies of apartheid called forth an escalation in the state's use of visible violence, to set in place a new regime of sovereign power that would last out the next 30 years.

Within the previous regime, violence had extended the eye of disciplinary surveillance into the depths of the mind and pervaded the spaces between bodies to mark out the interstices of tribe, culture, tradition, and inter-racial contact that were then made the targets of Native policy. Now, the new sovereignty meant that violence operated to curtail this individuating and illuminating tendency, installing in its place a great binary division between "the blacks" and "the whites". As a 1961 Pan Africanist Congress pamphlet put it:

We are starting again Africans … we die once. Africa will be free on 1 January. The white people shall suffer, the black people will rule. Freedom comes after bloodshed. Poqo has started. (cited in Reader's Digest, 1988, p. 411)

This, then, marked the beginning of a time of bombs, guns, bulldozers and brutal tortures, a time for which violence was only nominally a crime in violation of the law, any criminal justice pretensions to the contrary constantly contradicted by daily displays of state violence beating down on African bodies, and those very bodies fighting back with the self same weapons of sovereign power (see Karis & Carter, 1977, for an historical analysis of this swing to violence). The apartheid state thus operated in the first instance through overt political violence – conventional and counter-insurgency warfare, forced removals, assassinations, "disappearances", detention and torture – as well as through myriad forms of 'structural' violence. In response, the "armed struggle" began around the same time as the 1960 massacres at Sharpeville and Langa, and throughout the next two decades acts of violence increased exponentially, culminating in the dark years of the 1980s.

Under the weight of this oppression, the gaze of psychosocial surveillance that previously had pathologized the African mind as an object of white consciousness was reversed. In the 1970s, and as the single most prominent disciplinary counterpoint to the destructive power of sovereign violence, Biko's Fanon-inspired Black Consciousness (BC) emerged to invent an African personality which was a mirror image of that produced by ethnopsychiatry. "Essentially an inward looking process", the effect of BC was to expand the meaning of violence to include sociological and ideological factors, which it identified as destroying the authenticity of the black man and undermining the African's "pride and dignity". Hence the appeal of BC for the black man (sic) to:

… come to himself; to pump back life into his empty shell; to infuse him with pride and dignity; to remind him of his complicity in the crime of allowing himself to be misused and therefore letting evil reign supreme in the country of his birth. (Biko, 1988/1970, p. 43)

While ackowledging that not all oppressed persons were equally subject to the alienating effects of apartheid, the BC focus on how it insinuated itself into subjectivity meant that black persons themselves constituted the pathology, and therefore that their cure demanded rehabilitation of the entire individual and social body. This was emphasised by Manganyi's characterisation of the ordinary African as a "psychological paraplegic":

… in the African experience there was over time developed a sociological schema of the black body prescribed by white standards. The prescribed attributes of this sociological schema have, as we should know by now, been entirely negative. It should be considered natural under these circumstances for an individual black person to conceive of his body image as something entirely undesirable, something which paradoxically must be kept at a distance outside of one's self so to speak. (Manganyi, 1973, p. 51)

By problematising the sociological and ideological conduits through which apartheid oppression internalised itself in the minds and bodies of the oppressed, the effect of BC was to expand the variety of sites that were regarded as valid targets for violent resistance, and coinciding with its emergence violence spread beyond military installations and personnel to civilians and other non-military groups as the actual and perceived representatives of the apartheid state. Prominent among these non-military targets was the state system of "Bantu education" a pamphlet circulated to Soweto parents during the June 1976 protests against Afrikaans as a medium of education stating:

… you should rejoice for having given birth to this type of child … a child who prefers to die from a bullet rather than swallow a poisonous education which relegates him and his parents to a position of perpetual insubordination. (cited in Reader's Digest, 1988, p.444)

In 1977, a year after the Soweto protests, "Bantu" Steven Biko, the driving force of Black Consciousness, was himself murdered in an instance of political brutality that sensationally focused world attention upon the political use of violence by the South African state.

The frequency of politically-motivated attacks against the apartheid state and those seen as its members continued to increase after 1977, and these attacks enjoyed broad legitimacy from the majority of South Africans who saw no alternative, Posel writing that "from the early 1980s, mass resistance against apartheid erupted with unprecedented tenacity and ferocity" (1990, p.154). By the mid-1980s it was as if the long-established pattern of violence and counter-violence was spiralling out of control. Guerrilla attacks increased by 304 percent between 1984 and 1985 (Weekly Mail, 17-24, January, 1986 cited in Vogelman, 1995), and in 1986 and 1987 there were 230 and 240 guerrilla attacks respectively (Weekly Mail, 8-14, July, 1988 cited in Vogelman, 1995). As these attacks grew in frequency and state repression increased, so the idea of the armed struggle was increasingly popularised. The social acceptability of violence as a form of resistance was further boosted by the African National Congress (ANC) call to make townships ungovernable through any means possible (Vogelman, 1995), and by the declaration of 1986 as the year of the "people's army" (South African Institute of Race Relations, 1986).

Similarly, and as it had for many years, the state continued to claim legitimacy over its use of violent repression. Television viewers, for instance, were repeatedly told in 1986 that "we have a total revolutionary onslaught against us … we experience it every day" (cited in Posel, 1990, p.167). Such depictions of the state's integrity under siege were a way of legitimating its use of violence, for they allowed academics such as Van der Merwe (1989, cited in Hoffman & McKendrick, 1990) to argue that all nations validate and enforce violence in "the pursuit or protection of national interests" (p. 17), and National Party politicians to pontificate about the "total onslaught" and the threat of communism. These political and ideological positions found their tactical expression in the establishment of a "National Security Management System", which from the State Security Council downwards gave the military direct influence, through some 500 secret committees, over decision making down to the level of local government (Cock, 1989; 1990). This systematic "security" approach and the "low-intensity conflict" or counter-insurgency operations of the state exerted their pressure over the entire society, including the media and educational institutions, and soon polarised all into pro- and anti-apartheid camps.

By 1990, however, it was evident that the apartheid regime was collapsing under the combined force of internal destabilisation and external pressure (e.g. trade embargoes and other socio-economic sanctions). As democratic reform appeared imminent, so too were there attempts to deligitimate the use of violence for political ends. Thus, and amidst ongoing township violence at the inter-community level, the ANC denounced the "armed struggle", while the state (despite continuing to sponsor numerous "third force" actions) started moving towards the negotiated settlement that culminated in the historic April 1994 elections when sovereign power was formally passed from white to black hands. In post-apartheid South Africa the "legitimate" aspect of violence lingers in the form of sporadic flare-ups between ANC and Inkatha Freedom Party supporters in KwaZulu-Natal, which have resulted in 9 300 deaths and 61 reported massacres between 1990 and 1993 (Bronkhorst, 1995).

(iii) The new discipline I: caring for victims of violence

The re-emergence of sovereign power in South Africa during the apartheid years (as evidenced in spectacular displays of direct force by both oppressor and oppressed) constituted a hiatus in extension of the more "synaptic" forms of disciplinary power characteristic of the modern industrial nation. It was thus not only state-sanctioned racial discrimination, but also the crude methods of its enforcement that constituted a scandal to more progressive political sensibilities locally and internationally, and rather than respond solely with direct counter-force, many attempted to draw South African political practices back into a modernist, disciplinary configuration. Among such groups were the socio-medical sciences, which owe their very existence to the modernist power regime that has as its object and effect the individual. Consequently, mental health professionals attempted to counter the state's preoccupation with strategies of sovereign domination, and replace these with a progressive-humanist alternative that emphasised the individual, liberation, and empowerment, thrusting to prominence not the sources of violence but rather its victims.

Conditioned by this new regime of disciplinary power, the individual as the victim of violence emerged in two distinct but related forms. On the one hand, as a victim of ideological and structural violence who, viewed through the application by local scholars of Bulhan's (1985) interpretation of Fanon (1967), recalled the African personality that had been fabricated by BC nearly 20 years earlier. For instance, in attempting to explain the highly elevated rates of homicide and assault observed among blacks in comparison to whites, Nell and Brown wrote that:

Apartheid is a successor to colonialism, preserving the colonial structures of power and privilege. If Bulhan's hypothesis that 'vertical' institutional violence in colonial and neocolonial societies spreads horizontally among the victims of oppression is correct, it offers an explanation for the high homicide and assault rates among Africans. (Nell & Brown, 1991, p.295)

In comparison, the victim of violence as harmed by a traumatic event was by far the more prominent object of concern within this regime and the outcome of a somewhat different set of practices. This was the work done by (mainly white) psychiatrists, psychologists and medical practitioners, who in treating the victims of torture, detention, and injury resulting from police action (e.g. Perkel, 1990; Rayner, 1990; Solomons, 1988; Spitz, Eastwood & Verryn, 1990), risked decontextualizing the origins and outcomes of violence by reducing it to a discrete event with discrete consequences for the individual. Accordingly, this form of victimology was characterised by its own, self-reflexive critique, which tried to show that violence was an ongoing phenomenon (Straker, 1987); that detention "does not only produce negative psychological effects but is also a site and a source … of active human transformation towards a more just social order" (Foster & Skinner, 1990, p.229), and that the uncritical definition of persons as victims could be counterproductive (Swartz & Levett, 1989; Levett, 1989).

While certainly having somewhat different emphases, both these strands within the new discipline of the victim aimed to simultaneously illuminate the excesses and intensity of state violence, and to repair the damage this caused. Despite the extent of these confessional activities from the 1980s onwards, it was only in the mid-1990s – when the wider political arena started to shift back towards a disciplinary configuration – that their import began to filter upwards from the infra-policy level and into formal policy documents.

Early indications that this victim-oriented discourse was becoming formalised appeared in several draft documents produced by national and regional ANC health forums in the early 1990s. Within these documents mental health was continually tied to the socio-political context and violence began to be outlined as a mental health problem, requiring that those affected by it to be treated through systematic intervention programmes. These efforts to redraw violence as a disciplinary problem for the health and social sciences met, however, with little immediate success, the initial ANC Policy Guidelines (1992) for health omitting any mention of violence as a mental health concern. By 1994, however, violence was articulated as a key issue requiring urgent mental health attention. This was clearly documented in the ANC Health Plan (1994), the Reconstruction and Development Programme (1994), and subsequent policy documents including the Mental Health and Substance Abuse Committee Report (1995), and the Strategic Management Team task Group Gauteng Report on Mental Health (Zwi, Radebe, Ratemane, Freeman & Harris, 1995). Now, violence was constructed as a destructive heritage of past policies and practices that continued to exert a deleterious impact on communities and individuals. Additionally, these documents began to present a more fine-grained taxonomy of violence, discerning such acts as child abuse, women abuse, assault and other forms of criminal violence as discrete sub-categories. Their conclusion was that violence posed a major threat to the social and economic development of the country (Zwi et al., 1995), and it was argued that reconstruction and development were contingent on a safe society that assured physical and mental health. Where previously the curative goal had been to return victims of violence to their activist positions on the front line of the anti-apartheid struggle, now the focus was on restoring to victims of violence their capacity to serve as productive citizens.

Accordingly, the documents prioritised "victims of violence" for mental health interventions. Throughout, these victims were portrayed as psychologically wounded individuals in need of "adequate care" through the "establishment and support of crisis centres" and "counselling services" (ANC Health Plan, 1994; Zwi et al., 1995). Elsewhere, in a section that discussed policy frameworks for health care, the Reconstruction and Development Programme stated:

The RDP must aim to promote mental health and increase the quality, quantity and accessibility of mental health support and counselling services, particularly for those affected by domestic or other violence. (Reconstruction and Development Programme, 1994, 2.12.7.3)

Repeated mention was made of the need for specialised programmes for victims of violence, including victims of child abuse, women abuse and people at risk for violence owing to substance abuse. Complementing this curative approach within the mental health sphere was an equally reactive stance on the part of the criminal justice and policing sectors, which in emphasising a need for improved detection and punishment of violent perpetrators reproduced the post-event focus of mental health, and with it the idea that violence and its consequences could not be prevented, but only controlled.

(iv) The new discipline II: violence prevention

The burgeoning of victim-oriented practices in respect of violence marked the beginning of its translation from a mechanism of sovereign power into a conduit for the productive power of discipline. By the mid-1990s, this was being complemented by an increasing recognition that a curative approach was insufficient, and with it of the preventive possibilities set in place by epidemiological and risk-factor research into the causes of violence. Thus, while such research had already commenced in the late 1980s (eg. Nell & Butchart, 1989; Knottenbelt, 1989; Van der Spuy, 1989), it was only four to five years later that it began to attract the attention of the criminal justice, policing and health sectors as a key component in what was now a drive to make violence the object of a preventive regime that aimed to draw everyone into its web of surveillance.

This epidemiological renaissance extended the statistical surveillance of violence from the police perspective and into the health sector, and from the victim of violence as purely in need of care, to the victim of violence as at the same time a resource for information relevant to prevention. As this occurred, so hospital casualty wards, clinics and mortuaries began to serve as observatories for monitoring the demographic, personal and situational determinants of violence, alongside the earlier established human rights monitoring groups, police stations and state statistical services. This expansion in the sites from which violence could be scrutinised was to have a number of implications for formal policy. First, in extending the eye of epidemiology through the hospitals, it raised to visibility all those victims of violence which previously had not existed for a regime concerned only with high-profile cases of political violence, revealing (at least for Cape Town and Johannesburg) that of every 10 victims of violence treated in hospital emergency rooms, only one was injured in explicitly political conflict (Butchart et al, 1991b; Van der Spuy, 1993). Second, it problematised the perception that because violence was according to the law a criminal act, the best point from which to view it was through the police and court statistics of the criminal justice system. Now, it could be recognised that far from all violence entered into the criminal justice system, and that for many victims of violence (in South Africa as elsewhere), the health sector was the first and only point of contact, treating at least six times the number of victims that entered the criminal justice system (Butchart, Seedat & Nell, in press; National Committee for Injury Prevention and Control, 1989; Shepherd, Shopland, Pearce & Scully, 1990).

Although these recognitions coincided with the emergence of a number of non-governmental organisations that applied epidemiological insights in community based approaches to violence prevention (for a review, see Butchart, 1996), they did not find rapid translation into formal policy. While the 1994 ANC Health Plan noted that only 15 percent of violence was political, it failed to elaborate on where the causes of the remaining 85 percent might lie, or to develop the concept of violence prevention beyond noting the need for preventive programmes. Similarly, the Gauteng Strategic Management Team Report (Zwi et al., 1995) highlighted violence as an important health care problem, but discussed it under secondary and tertiary modalities of care only, underlining the need for more staff to cope with the mental health problems associated with violence and the development of mobile trauma counsellors, but failing to mention violence at the primary health care level. All the policy documents studied failed to develop a comprehensive violence prevention programme which offered a broad analysis of violence incorporating domestic, criminal and community development issues.

Despite these failings, the documents did convey an implicit sense of violence as a preventable problem. By continually linking the state of health care to the social context, improvements in all aspects of social life (e.g. housing, employment and social development) were presented as prerequisites for the prevention of violence and promotion of a mentally "healthy" society. There was an emphasis on restructuring various institutions (e.g. security forces, correctional services) that in the past were instrumental in the creation and maintenance of systematic violence, and through this to cultivate the growth of a human rights culture that would discourage violence (cf. Mental Health & Substance Abuse Committee, 1995; RDP, 1994, Zwi et al., 1995). Cognitively, at least, violence had by 1995 been radically transformed from a political and criminal issue into a sociological and public health problem, a collective pathology that originated not merely from the formal centres of political power and resistance, but which was caused by the very fabric of society of which it could now be seen as an integral component – the violence in and of everyday life.

As a result, the new discourse defined violence as an intersectoral problem, a challenge not for the police, criminal justice, educational or health sectors acting independently, but one which could only be solved through their joint efforts, and, most importantly, through recruitment of the very individuals and communities most affected by it into the task of prevention and control. For instance, 1994 ANC health guidelines advocated a single mental health dispensation. Founded on a Primary Health Care approach, this emphasised collaboration, community participation, community orientation of services, empowerment of individuals and cost-effectiveness. Preventive, promotive, curative and rehabilitative forms of service were to be provided alongside each other on a rational and cost-effectiveness basis (ANC, 1994; RDP, 1994; Mental Health & Substance Abuse Committee, 1995; RDP, 1994, Zwi et al., 1995), and violence was isolated as one of the most pressing problems requiring the involvement of more than one ministry (cf. Mental Health & Substance Abuse Committee, 1995; RDP, 1994, Zwi et al., 1995). Intersectoral planning committees were envisaged as being made up of government departments, non-governmental organisations, consumers, and other mental health care role players, the chief responsibility of these bodies being to enable co-operation and ensure that fragmentation and duplication of services were minimised (Zwi et al., 1995). In May 1996, this vision found further elaboration in the National Crime Prevention Strategy document (National Crime Prevention Strategy, 1996), which identified violence as among the top priorities for crime prevention, attributed its occurrence to a broad range of political, social, economic and environmental causes, and stated that the fundamental challenge was to establish a multi-agency preventive task force.

The relationship between crime, violence and development necessitates the engagement of the Reconstruction and Development Programme and of developmental agencies. The imperative of delivering comprehensive victim empowerment strategies and services implies the involvement of the Social Welfare and Health Departments. The impact of economic deregulation and the need for some forms of regulation to control criminal markets appears to require the participation of the Department of Trade and Industry in any crime prevention strategies. The problems of regional security and illegal immigration necessitate the focused attention of the SANDF, SASS and the Department of Home Affairs respectively. (National Crime Prevention Strategy, 1996, p.44)

However, as of mid-1996, this rhetorical commitment to intersectoriality has yet to be matched by any equivalent shift in practice, other than sporadic attempts by non-governmental organisations to engage with violence prevention stake holders from the state, the community, and other sectors. For example, while the Truth and Reconciliation Commission provides a unique opportunity for intersectoral work between the departments of Justice, Welfare, Health, and the Directorate of Mental Health, and despite lobbying from various non-governmental groupings (reviewed in Hamber, 1995a, Hamber, 1995b), no formal involvement by any ministry other than Justice had occurred by mid-1996.

Possible reasons for the failure of this intersectoral vision of violence prevention to translate into action are many, and, perhaps most importantly, include the apartheid-imposed break in the theory and practice of social medicine (for reviews, see Butchart, 1995; Tollman, 1991; Yach & Tollman, 1993). From the late 1950s to the early 1990s, this hiatus in public health practice meant that locally there was no clear model of how to achieve transectoral involvement, and no organisational structure through which to deliver multi-agency services in respect of information systems, programme design, service delivery and evaluation. However, with the recent establishment of public health schools in the western Cape and Gauteng, and with increasing involvement in South Africa by global public health players such as the World Health Organisation (which at its May 1996 World Assembly adopted a resolution prioritising violence as a public health problem), it is reasonable to believe that an interdisciplinary public health policy for violence prevention may now have some chance of finding successful implementation.

A Public Health Policy for Violence Prevention

Against the traditional criminal justice idea of violence control through retribution and incapacitation, and the curative orientation of mental health and clinical medicine, the public health approach reflects a fundamental shift. Where the traditional approaches individualise the problem by locating the causes and consequences of violence within the person, the public health approach recognises the neurological, physiological and psychological make up and behaviour of individuals as the outcome of environmental and socialisation factors operating at the level of the community and society. Where the traditional approach is limited to reacting to violence through post-hoc attempts at cure and control, the public health emphasis is on preventing violence before it occurs. Where the traditional approach places the primary burden of control on the 'police' and the hospital, the public health approach acknowledges that the causes of violence can only be addressed through inter-disciplinary and cross-sectoral involvement.

In short, the public health approach locates violence at the centre of the three overlapping perspectives provided by the criminal justice system, the social sciences, and the disciplines of public health. For the criminal justice system, violence is a violation of the law. For the social sciences, it is a form of aggressive behaviour with harmful consequences for the individual, family and community, while for public health it is the cost in terms of death and disability that forms the primary motivation for involvement. Unifying this approach is the public health emphasis on the importance of science. This is the acknowledgement that "effective policies for preventing violence must be firmly grounded in science and attentive to unique community perceptions and conditions. Scientific research provides information essential to developing such policies and prevention strategies and methods for testing their effectiveness" (Mercy et al, 1993, p. 8).

The promise of the public health approach to generate new insights into violence prevention emanates from the success it has achieved in the prevention of occupational and motor vehicle injuries, through which the so-called "Haddon matrix" was developed (Haddon, 1980). This simple though effective aid to analysis, strategy identification, planning, and resource allocation consists in its most basic form in a matrix of two dimensions. The first dimension recognises that all violent injuries are events located in time and therefore that their damaging effects are preceded by processes that divide naturally into three stages: Pre-event, event, and post-event. The second dimension derives from the epidemiological triangle of disease analysis, and recognises that all injuries are determined by three broad classes of factors. First, those that shape the susceptibility or resistance to violence of the potential victim, such as age, verbal ability, physical strength and childhood experiences. Second, those that pertain to the agent of violent injury, usually subdivided into "perpetrator" and "weapon". Third, all those environmental factors, both physical and socio-cultural, that increase or reduce the person's threshold for violent behaviour – such as the degree of defensible space allowed by the built environment, and socio-economic deprivation.

A source of the public health model's innovation is its insertion of a multi-disciplinary scientific approach into the field of violence prevention, which operates in four stages:

The first stage of the model involves problem definition, the second identification of risk factors and causes, the third development and evaluation of interventions, and the fourth intervention implementation in sustained prevention programmes. While described as a linear progression from problem definition to response, in practice these steps are likely to occur simultaneously, information systems serving as key elements in the monitoring and evaluation of programmes, and insights obtained through programme implementation suggesting new types of intervention.

(i) Problem definition. For the public health approach, violence quite simply has no existence as a problem amenable to management unless it is subject to repeated and ideally ongoing epidemiological study. Furthermore, problem definition from the public health perspective demands that violence be analysed in relational terms so as to reveal sites for intervention in the lines between persons, products and environmental factors that sustain violence. Accordingly, as a first step, it is recommended that data collection efforts must go beyond the mere enumeration of cases to illuminate the demographic characteristics of victims and perpetrators, the victim-perpetrator relationship, the geographical and temporal characteristics of incidents, the mechanisms of injury, the involvement of alcohol, and so on. Such information will permit the design of appropriate interventions keyed to discrete sub-types of violence. For instance, that most female victims of violence are attacked in the privacy of the home with sharp weapons by men they know intimately, demands a far different approach than attempts to prevent firearm-related violence that occurs among strangers on the street.

Problem definition typically proceeds through the collection of data using time-bound epidemiological studies, and methods of continuous, ongoing injury surveillance. For non-fatal injuries, hospital casualty departments and admission wards are the most obvious site for problem definition (e.g. Brown & Nell, 1991; Butchart et al, 1991a; National Trauma Research Programme, 1994). This is due both to convenience, and the fact that these sites show a substantially larger portion of the violence iceberg than do the statistics on violent incidents reported to the police. In South Africa, completed epidemiological studies of violence have focused only on Cape Town and Johannesburg (Butchart et al, 1991, a,b; Van der Spuy, 1993), and are now of largely historical value only, having been conducted in 1990 and therefore unlikely to reflect the current situation as it has evolved since the relaxation of apartheid laws governing place of residence and regulating migration to large urban areas. There are however, some studies currently under way, most notably a hospital and mortuary-based survey by the Medical Research Council in Kwa-Zulu Natal, and an extensive house to house survey of violence and injury in a Johannesburg community composed of informal settlements, council housing, and council flats (Butchart, Kruger, Lekoba & Smith, in press). Rather than regarding these initiatives as sufficient in themselves, the proper policy response should be to examine them for the lessons they may offer in researching the causes of violence in other geographical areas, and using their findings to improve the design of data collection forms for use in sites where prevention information can be collected on an ongoing basis, such as police stations and hospitals.

Against time-bound epidemiological studies, the surveillance of violent injuries involves the collection, analysis, and reporting of appropriate data on a constant and ongoing basis. For non-fatal injuries, this once again is most readily achieved through the use of hospital data. For fatal violence, the surveillance sites of choice are the state mortuaries (see Lerer, 1993a). Not all injury deaths seen in the mortuaries reach the inquest court, and because deaths due to injury are regarded as non-natural in terms of current legislation, a statutory requirement exists for the performance of medico-legal autopsy examinations on all victims of violent and unintentional injury. Besides the post-mortem examination itself, various other examinations such as blood alcohol levels and toxicological analysis may be performed, and these, together with the actual report on the circumstances surrounding death (Lerer, 1993b), can provide valuable information in the area of violence prevention. To date, and due to the enterprise of a particular individual rather than any systematic policy, only Cape Town maintains a fatal injury surveillance programme (Lerer, Matzopolous & Bradshaw, 1995). This has already directly impacted upon violence prevention through its success in pressurising SA Metro to initiate interventions on its commuter rail lines, where a combination of intentional and unintentional injuries kill on average one person every two days (Lerer & Matzopolous, 1995).

While there is little reason to doubt that South Africa is among the world's more violent countries, there is at the same time every reason to be concerned about why the problem continues to be so inadequately defined. For only if high quality data about the determinants and distribution of violence are made available on a routine and ongoing basis to a broad coalition of policy makers and intervention agencies can we even begin to set in place and evaluate prevention programmes.

(ii) Risk factor identification. Problem definition, while a necessary component, does not in itself provide a sufficient basis for prevention. Thus, while it looks at the how, when, where and what of violence, the second step of the public health approach – risk factor identification – looks at why. Why, for instance the marked intra-urban differentials in the distribution of injury causes between equally impoverished areas (e.g. Dorrenboom, 1995), or the clear relationship between age, gender and risk for violence? To answer such questions demands the deployment of experimental and qualitative research designs that go beyond identifying risk in terms of largely unmodifiable characteristics such as age or gender. For instance, while everyone knows that young black and "coloured" men aged 15 to 24 are most at risk for violence, this says nothing about what differentiates those members of this cohort who do become victims or perpetrators from those who do not. Clearly, if these factors can be identified through case-control and cohort studies – as Farrington has done for high risk groups in the UK (e.g. Farrington, 1989) – then important data for social prevention programmes that target prenatal, perinatal and pre-adolescent developmental factors will be made available. So too with the problem of alcohol. Local studies show that some 50 to 70 percent of all victims of violence are intoxicated at the time of attack (Butchart & Brown, 1991; Lerer, 1992, Van der Spuy, 1993), but are silent on the factors that sustain this close relationship between violence and alcohol use. Perhaps, as Butchart, Lerer and Terre Blanche (1994) have shown, it is in part related to the popular stereotype – perpetuated by aggressive marketing – of alcohol as a benign intoxicant. Conducted in Cape Town, this study showed that while 52 percent of all female homicide victims seen in 1992 had elevated blood alcohol levels, only one out of 50 women asked to imagine their own violent deaths mentioned alcohol use as a precipitating factor.

While it is perhaps true that many risk factors identified by studies elsewhere can be generalised to local conditions, this does not dispel the need for similar research here. Unfortunately, as is the case for problem definition, local studies of violence risk factors are all but non existent (but see Butchart, Lerer & Terre Blanche, 1994; Flisher et al, 1993; Lerer, 1992), highlighting the immense gap that exists in our knowledge of how an individual enters the path of violence and precisely what combination of factors render one interpersonal situation violent and another comparable one non-violent.

(iii) Develop and test interventions. Through problem definition and risk factor analysis, violence is rendered manageable by casting it in terms of relational factors and revealing the lines between people, products, and the environment that sustain contexts of violence. The third step in the public health approach to violence prevention is therefore to use these lines of risk to design and implement interventions.

Although interventions should always be calibrated to the perceived and externally identified needs and risk factors at the level of the community, they can at the abstract level be classified into: (a) educational interventions aimed at changing individual skills, knowledge and attitudes; (b) engineering aimed at the reduction of risk by improving the safety of dangerous products and places (e.g. low-cost housing designed with explicit attention to building out violence; training of alcohol servers to intervene where violence threatens; the design of commuter rail carriages), and (c) the establishment and enforcement of legal deterrents against the abuse of products that increase the risk for violence (e.g. guns, alcohol, drugs). A further important distinction is between interventions according to the level of activity demanded by the individual or group targeted. Active interventions are those that require much action on the part of the individual, while passive interventions concern those measures that require no individual action. It has been consistently shown in relation to a range of public health problems that because they do not require the sustained and active cooperation of the public, passive prevention strategies are generally more efficacious than active measures. Accordingly, and where feasible, a much higher value and priority should be placed on passive over active measures (Haddon, 1980, p.416).

In relation to violence, a number of passive measures suggest themselves, such as legislation governing the design of low-cost housing and lay-out of estates, the fitment of metal detectors in public spaces, the strategic illumination of high-risk areas, and the control of alcohol use through pricing. While it is clear that such interventions can only take hold through top-down procedures that make them into law, the impetus for installation should wherever possible be harnessed to bottom-up, community based initiatives, in that way ensuring a good fit between legislation, environmental modification and popular perceptions. In practice, therefore, violence prevention initiatives are most commonly community-based and involve a mix of educational, engineering and enforcement components (see Cohen & Lang, 1990; National Center for Injury Prevention and Control, 1993).

(iii) Implement interventions. The final step of the public health model involves the implementation of interventions that have been proven effective or are highly likely to prove effective. It is here that successful pilot programmes are translated out of their experimental phase and with state or provincial support inscribed in public health policy. As of mid-1996, nowhere in South Africa had this stage been reached, and until sufficiently compelling proof for the efficacy of the public health approach becomes available, it will not be reached. However, this is not the same as saying that the public health approach should be marginalised until it proves itself. On the contrary, unless there is solid provincial or state support for injury surveillance and risk factor research by which pertinent data are regularly collected and disseminated, the point at which the effectiveness of the public health approach – or, indeed, any other strategy – is proven will never be reached, since it is precisely such information systems that make possible the outcome evaluation of violence prevention programmes.

The public health approach to violence is a novel one, not only to South Africa but also nationally and internationally. Empirical evidence for its efficacy in preventing violence is, however, beginning to emerge from other countries. Most promising are early developmental interventions (e.g. Berrueta-Clement et al, 1984; Farrington, 1993; Kazdin, Siegel & Bass, 1992), which target high-risk infants and their mothers through a combination of home-visiting, pre-school training, and parent effectiveness training. Some success has also been shown in respect of schools-based programmes that train youth to identify potentially violent situations and manage interpersonal confrontations using pro-social alternatives to violence (Gainer, Webster & Champion, 1993; Laird, Syropoulos & Black, 1993; Northrop, Jacklin, Cohen & Wilson-Brewer, 1990). Lastly, gun violence through regulatory interventions that enforce laws against weapon carrying in public have also returned success (e.g. McDowall, Loftin & Wiersema, 1992). These examples of success are all from the US, and as yet there is no scientific proof that the public health approach to violence prevention can be successfully adapted the developing-country context of South Africa.

Lack of evidence for its efficacy is, however, no cause to rule the public health approach out of court, for the self-same public health principles have met with remarkable success in the prevention of car crashes, home and occupational injuries, smoking and many infectious diseases. Thus, until properly coordinated and sustained efforts are made to apply and evaluate its application to violence prevention in South Africa, we quite simply have no evidence by which to judge its utility. To conclude, we should therefore take what can be learned from the present deficits in our society's response to violence to set out a series of violence prevention priorities. In keeping with the four-steps of the public health model, these are as follows.

1. Surveillance. Top priority must be given to policy development that will mandate for the development and implementation of a uniform system to collect etiological data on fatal and non-fatal injuries from state mortuaries and hospitals. Owing to the relative simplicity of data collection from mortuaries as opposed to hospitals, fatality surveillance should receive immediate attention from an expert task group comprised of forensic epidemiologists, social scientists and police detectives. Such a surveillance system should be open to regular audit by outside agencies to ensure accuracy of coding, and data analysis should be performed by appropriate scientists. Detailed information should be regularly disseminated at the local level of each magisterial district where it can be used to design proactive and community-specific interventions, while summary data should be communicated to the Central Statistical Services for the monitoring of national trends. Similar considerations should apply to the more complex problem of hospital-based surveillance for non-fatal injuries.

2. Violence and injury control centres. Provincial and national centres for violence and injury control should be established with the goal of creating committed and constructive interdisciplinary and cross-sectoral engagement that will concentrate and systematize the currently dispersed and therefore disempowered and unmeasurable efforts of the many individuals, groups and agencies already committed to violence prevention.

3. Risk factor research. While more research into all aspects of violence are urgently required, three key areas demand immediate attention if prevention is to proceed. First, the relationship between alcohol and violence. Second, more precise delineation of the risks and benefits attaching to the possession of and ready access to firearms (eg Kellerman et al, 1993). Third, research that will articulate the direct and indirect costs to the economy resulting from violent death and disability, as a basis both for lobbying to have increased resources allocated for violence prevention activities, and to permit cost-benefit analyses of various prevention strategies.

   The remarkably impoverished state of our present response to violence means that any number of further recommendations could be made, such as the calculation and dissemination of police clearance rates at community level, the training of more public health professionals, and the need for international collaboration around the disablement of illegal weapon supply networks. Many of these issues are doubtless addressed by the more police- and criminal justice-oriented approaches to the management of violence. Suffice to say here that if such interventions can be brought into closer collaboration with the public health contribution to violence prevention, then we will be well on the way to establishing a truly coordinated and therefore viable response to violence in our society, our communities and our homes.

Conclusion

In common with most other contemporary understandings of violence, this chapter has argued that violence is the outcome of a multitude of factors, not least of which are the particular ways in which it is socially constructed in the course of different discursive regimes. As South Africa in the twentieth century has oscillated between sovereign and disciplinary modes of knowledge production, so mental health, as the quintessential disciplinary enterprise, has contributed in varying degree to what is taken to be the true nature of violence, its causes and possible cures. At certain points, such as during the era of the 'African personality', and again during the era of the African as victim of apartheid, classical mental health conceptions were central to delineating the origins and effects of violence, while at other times violence appeared to drift beyond the reach of discipline and into the hands of sovereignty.

The resurgence of civil society in the late 1990s again positions violence clearly as a disciplinary problem, and therefore a challenge to the mental health and the socio-medical sciences more generally. We maintain that the public health approach, as it is being extended to behavioural and mental health problems, provides the most likely way forward, and that if it is to maximise its collaborative contribution the mental health sector must find innovative ways of inserting its insights regarding violence into and through the many authorities and agencies that on a daily basis interact with actual and potential perpetrators and victims. For, while it certainly is important that mental health professionals continue to be direct service providers, they are in many instances less well-positioned to interact with individuals and groups at risk for violence than are the teachers, police, prison warders, doctors and nurses, bar and shebeen owners, community agencies and other bodies that on a day-to-day basis interact with people who have been violated or may themselves be violent. In practice, this will mean taking a step back from the front line of service provision, and, instead, developing the capacity of such agencies to deliver appropriate psychosocial responses and install proactive prevention measures in each and every interstice of the social body and individual subjectivities that through their practices they continuously invent, sustain and transform.

References

Albertyn, J.R. (1932). The poor white problem in South Africa. Report of the Carnegie Commission, Part V. Sociological report. Stellenbosch: Pro-Ecclesia.

African National Congress. (1994). A National Health Plan for South Africa. African National Congress, May 1994.

African National Congress. (1992). Ready to Govern: ANC Policy Guidelines for a Democratic South Africa. Adopted at the National Conference, May 1992. Published by the Policy Unit of the African National Congress.

Berrueta-Clement, J.R., Schwein hart, L.J., Barnett, W.S., Epstein, A.S., & Weikart, D.P. (1984). Changed lives. Ypsilanti, Mich.: High/Scope.

Biesheuvel, S. (1953). A technique for measuring attitudes of educated Africans. Proceedings of the South African Psychological Association, 4, 13-20.

Biesheuvel, S. (1955). The measurement of African attitudes toward European ethical concepts, customs, laws and administration of justice. Journal of the National Institute for Personnel Research, 6(1), 5-17.

Biesheuvel, S. (1957). The influence of social circumstances on the attitudes of educated Africans. South African Journal of Science, 53, 309-314.

Biko, S. (1988/1970). We Blacks. In A. Stubbs (ed.) I write what I like (pp.41-46). Harmondsworth: Penguin Books.

Bronkhorst, D. (1995). Truth and Reconciliation: Obstacles and Opportunities for Human Rights. Amsterdam: Amnesty International Dutch Section.

Brown, D.S.O., Nell, V. (1991). Epidemiology of traumatic brain injury in Johannesburg – I. Methodological issues in a developing country context. Social Science and Medicine, 33, 283-287.

Bulhan, H.A. (1990). Afrocentric psychology: Perspectives and practice. In L.J. Nicholas and S. Cooper (eds.) Psychology and apartheid (pp.67-78). Johannesburg: Vision Publications.

Bulhan, H.A. (1985). Frantz Fanon and the psychology of oppression. New York: Plenum.

Butchart, A. (1996). Injury prevention policy in South Africa: What can be learned from current violence prevention programmes? Paper presented at the Third International Injury Prevention and Control Conference. Melbourne: 18-22 February.

Butchart, A. (1995). On the anatomy of power: Bodies of knowledge in South Africa socio-medical discourse. Unpublished D Litt et Phil thesis. Pretoria: University of South Africa.

Butchart, A., Brown, D.S.O. (1991). Non-fatal injuries due to interpersonal violence in Johannesburg-Soweto: Incidence, determinants and consequences. Forensic Science International, 53, 35-51.

Butchart, A., Kruger, J., Lekoba, R. & Smith, D. Evaluating injury prevention outcome in a developing country context: Lessons from a community-based violence prevention programme. In press, Urbanisation and Health Newsletter.

Butchart, A., Lerer, L B., Terre Blanche, M. (1994). Imaginary constructions and forensic reconstructions of fatal violence against women: Implications for community violence prevention. Forensic Science International, 64, 21-34.

Butchart, A., Nell, V., Yach, D. et al. (1991a). Epidemiology of non-fatal injuries due to external causes in Johannesburg-Soweto. Part I. Methodology and materials. South African Medical Journal, 79, 466-471.

Butchart, A., Nell, V. Yach, d. et al. (1991b). Epidemiology of non-fatal injuries due to external causes in Johannesburg-Soweto. Part II. Incidence and determinants. South African Medical Journal, 79, 472-479.

Butchart, A., Seedat, M. & Nell, V. Violence in South Africa: Its definition and prevention as a public health problem. In press in J. Seager and C. Parry (Eds.) Urbanisation and Health in South Africa. Cape Town: Medical Research Council.

Cohen, S. & Lang, C. (1990). Application of principles of community-based programmes. Background paper prepared for the Forum on Youth Violence and Minority Communities: Setting the Agenda for Prevention. Atlanta GA: Centers for Disease Control and Prevention.

Cock, J. (1989). Introduction. In J. Cock & L. Nathan (Eds), War and Society: The Militarisation of South Africa. Cape Town: David Philip.

Cock, J. (1990). Political Violence. In B.W. McKendrick and W.C. Hoffman (Eds), People and Violence in South Africa. Cape Town: Oxford University Press.

De Kock, V. (1950). Those in Bondage. London: George Allen and Unwin.

De Ridder, J. (1961). The personality of the urban African in South Africa (a thematic apperception test study). London: Routledge & Kegan Paul.

Dorrenboom, J. (1995). Colours that clash. Unpublished Master's dissertation. Leiden: University of Leiden.

Dubow, S. (1987). Race, civilization and culture: The elaboration of segregationist discourse in the inter-war years. In S. Marks and S. Trapido (Eds.), The politics of race, class and nationalism in twentieth century South Africa (pp.71-94). Harlow: Longman.

Editorial. (1912). An urgent question of sociology. South African Medical Record, 10, 201-203.

Editorial. (1968). Violence. South African Medical Journal, 42 (9), 737.

Erasmus, P.F. (1975). Thematic Apperception Test (TAT-Z). Prudery: Institute for Psychological and Audiometric Research, Human Sciences Research Council.

Erasmus, P.F. (1984). Manual for the Thematic Apperception Test (Zulu) (TAT-Z). Prudery: Human Sciences Research Council.

Fanon, F. (1967). Black skin, white masks. New York: Grove Press.

Farrington, D.P. (1993). Success stories in the prevention of adolescent aggression and youth violence. Plenary presentation to the Second World Injury Prevention and Control Conference. Atlanta G.A.: March.

Farrington, D.P. (1989). Early predictors of adolescent aggression and adult violence. Violence and Victims, 4, 79-99.

Flisher, A.J., Ziervogel, C.F., Chalton, D.O., Leger, P.H., Robertson, B.A. (1993). Risk taking behaviour of Cape Peninsula High School students. VII. Violent behaviour. South African Medical Journal, 83, 490-494.

Foster, D. (1991). Historical and legal traces, 1800-1900. In S.Lea and D.Foster (Eds) Perspectives on mental handicap in South Africa. Durban: Butterworths.

Foster, D. & Skinner, D. (1990). Detention and violence: Beyond victimology. In N.C. Manganyi and A du Toit (Eds.), Political violence and the struggle in South Africa (pp.205-233). Halfway House: Southern Book Publishers.

Foucault, M. (1977). Discipline and punish. New York: Pantheon Books.

Foucault, M. (1979). The history of sexuality, Vol. 1. London: Allen Lane.

Freeman, M. (1992). Providing Mental Health Care for All in South Africa. Structure and Strategy. Johannesburg, University of the Witwatersrand: Centre for Health Policy.

Gainer, P.S., Webster, D.W. & Champion, H.R. (1993). A youth violence prevention programme. Archives of Surgery, 128, 303-308.

Gordon, C. (Ed.). (1980). Power/Knowledge: Selected interviews and other writings 1972-1977 by Michel Foucault. Brighton: Harvester Press.

Haddon, W. (1980). Advances in the epidemiology of injuries as a basis for public policy. Public Health Reports, 95, 411-421.

Hamber, B. (1995a). Dealing with the Past and the Psychology of Reconciliation. The Truth and Reconciliation Commission: A Psychological Perspective. Paper presented at the 4th International Symposium on the Contributions of Psychology to Peace. Cape Town: June, 1995

Hamber, B. (1995b). Do Sleeping Dogs Lie? The Psychological Implications of the Truth and Reconciliation Commission in South Africa. Johannesburg: Centre for the Study of Violence and Reconciliation.

Hoffman, W.C. & McKendrick, B.W. (1990). The Nature of Violence. In B.W. McKendrick and W.C. Hoffman (Eds), People and Violence in South Africa. Cape Town: Oxford University Press.

Kazdin, A.E., Siegel, T.C. & Bass, D. (1993). Cognitive problem-solving skills training and parent management training in the treatment of antisocial behaviour in children. Journal of Clinical and Consulting Psychology, 60, 733-747.

Karis, T. & Carter, G.M. (Eds.). (1977). From protest to challenge: A documentary history of African politics in South Africa 1882-1964. Vol. 3 – Challenge and violence 1953-1964. Hoover Institution Press.

Kellerman, A., Rivara, F.P., Rushforth, N.B. et al. (1993). Gun ownership as a risk factor for homicide in the home. New England Journal of Medicine, 329, 1084-1091.

Knottenbelt, J.D. (1989). Trauma, the scourge of modern society (editorial). South African Medical Journal, 75, 1-2.

Kritzman, L.D. (Ed.). (1988). Michel Foucault: Politics, philosophy, culture: Interviews and other writings, 1977-1984. New York: Routledge.

Laird, M., Syropoulos, M. & Black, S. (1996). Aggression and violence: The challenge for Detroit schools. Newark: Lions-Quest International.

Laubscher, B.J.J. (1935). Sex, custom and psychopathology (a study of South African pagan natives). London: George Routledge and Sons.

Lerer, L.B. (1992). Women, homicide and alcohol in Cape Town, South Africa. Forensic Science International, 53, 93-99.

Lerer, L.B. (1993a). "Morti vivos docent": What death teaches us about injury. Trauma Review, 2(1): 7.

Lerer, L.B. (1993b). Improving mortality data in South Africa: Review of next of kin statements to determine cause of death in police certification. Journal of Epidemiology and Community Health, 43, 248-254.

Lerer, L.B. & Matzopolous, R.G. (1995). Meeting the challenge of rail commuter injuries: The public health approach (technical report, community health research group). Cape Town: Medical Research Council.

Lerer, L.B., Matzopolous, R. & Bradshaw, D. (1995). A profile of non-natural mortality in the Cape Town metropole 1994. Cape Town: Medical Research Council.

Levett, A. (1989). Psychological trauma and childhood. Psychology in Society, (12), 19-32.

MacCrone, I.D. (1957). Race attitudes in South Africa: Historical, experimental and psychological studies. Johannesburg: Witwatersrand University Press.

Manganyi, N.C. (1973). Being-Black-in-the-world. Johannesburg: Ravan Press.

Manganyi, N.C. (1972). Body image boundary differentiation and self-steering behaviour in African paraplegics. Journal of Personality Assessment, 36(1), 45-50.

McCulloch, J. (1995). Colonial psychiatry and the African mind. Cambridge: Cambridge University Press.

McDowell, D., Loftin, C. & Wiersma, B. (1992). A comparative study of the preventive effects of mandatory sentencing laws for gun crimes. The Journal of Criminal Law and Criminology, 83, 378-394.

Mental Health & Substance Abuse Committee (1995). Report on Mental Health and Substance Abuse. Unpublished policy document submitted to the Department of Health, Gauteng

Mercy, J.A., Rosenberg, M.L., Powell, K.E., Broome, C.V. & Roper, W.L. (1993). Public policy for violence prevention. Health Affairs, Winter Edition, 7-29.

National Committee for Injury Prevention and Control. (1989). Injury Prevention: Meeting the Challenge. American Journal of Preventive Medicine, 5(3): supplement.

National Crime Prevention Strategy Team. (1996). National crime prevention strategy. Unpublished typescript, dated May 1996. Place of publication unlisted: Various SA Government Departments.

National Trauma Research Programme. (1994). A guide to rural injury data capture. Cape Town: Medical Research Council.

Nell, V., Brown, D.S.O. (1991). Epidemiology of traumatic brain injury in Johannesburg – II. Morbidity, mortality and etiology. Social Science and Medicine, 33, 289-296.

Nell, V. & Butchart, A. (1989). Studying violence in a South African city. Critical Health, 28, 44-49.

Northrop, D., Jacklin, B., Cohen, S., Wilson-Brewer, R. Violence prevention strategies targeted towards high-risk minority youth. US Department of Health and Human Services and Centers for Disease Control and Prevention. 1990.

Perkel, A. (1990). Psychotherapy with detainees: A theoretical basis. Psychology in Society, (13), 4-16

Phillips, R.E. (1948). The Bantu in the city. A study of cultural adjustment on the Witwatersrand. Lovedale: The Lovedale Press.

Posel, D. (1990). Symbolizing violence: State and media discourse in television coverage of township protest, 1985-7. In N.C. Manganyi and A. Du Toit (Eds.). Political violence and the struggle in South Africa (pp.154-171). Halfway House: Southern Books.

Rayner, M. (1990). From Biko to Wendy Orr: The problem of medical accountability in contexts of political violence and torture. In N.C. Manganyi and A. Du Toit (Eds.). Political violence and the struggle in South Africa (pp.172-204). Halfway House: Southern Books.

Readers' Digest. (1988). Illustrated history of South Africa. New York: The Reader's Digest Association.

Reconstruction and Development Programme (1994). The Reconstruction and Development Programme: A Policy Framework. African National Congress.

Sachs, W. (1937). Black hamlet: The mind of an African negro revealed by psychoanalysis. London: Geoffrey Bles.

Shepherd, T.P., Shopland, M., Pearce, N.X., & Scully, C. (1990). Pattern, severity and aetiology of injuries in victims of assault. Journal of the Royal Society of Medicine, 83, 75-78.

Sherwood, E.T. (1957). On the designing of TAT pictures, with special reference to a set for an African people assimilating western culture. Journal of Social Psychology, 45, 161-190.

Silove, D. (1990). Doctors and the state: Lessons from the Biko case. Social Science and Medicine, 30, 417-429.

Solomons, K. (1988). A contribution to a theory of the dynamic mechanisms of Post-traumatic Stress Disorder in South African detainees. Psychology in Society, (11), 18-30.

South African Institute of Race Relations. (1986). Survey of race relations in South Africa, 1986. Johannesburg: South African Institute for Race Relations.

Spitz, S., Eastwood, R. & Verryn, P. (1990). The torture goes on: The psychology of restrictions. Psychology in Society, (13), 17-26

Straker, G. 1987. The continuous traumatic stress syndrome – the single therapeutic interview. Psychology in Society, (8), 46-79

Swartz, L., & Levett, A. (1989). Political repression and children in South Africa: The social construction of damaging effects. Social Science and Medicine, 28, 741-750.

Transvaal Archives Depot. (202/10 LD 1786). Dr. J.G. Visser: Re liberation of native lunatics from the lunatic asylum. February-March, 1910.

Transvaal Archives Depot. (CS 863/14966). Lunacy Amendment Act, 1908.

Van der Spuy, J. (1993). Violence in perspective. Trauma Review; 1(2): 1-4.

Van der Spuy, J. (1989). More about trauma. South African Journal of Continuing Medical Education, 7, 721-726.

Van Onselen, C. (1982). Studies in the social and economic history of the Witwatersrand. Vol. 2. New Nineveh. Johannesburg: Ravan Press.

Vaughan, M. (1991). Curing their ills: Colonial power and African illness. Cambridge: Polity Press.

Vogelman, L. (1995). The Pathway to Murder: A Social Psychological Study of the Evolution of Violence in an Industrial Dispute. Unpublished Ph.D Thesis, University of London, Department of Organisational Psychology, Faculty of Science, Birkbeck College, London.

Yach, D. & Tollman, S. (1993). Public health initiatives in South Africa in the 1940s and 1950s: Lessons for a post-apartheid era. American Journal of Public Health, 83, 1043-1050.

Zwi, A.B. (1987). The political abuse of medicine and the challenge of opposing it. Social Science and Medicine, 25, 649-657.

Zwi, A.B. & Ugalde, A. (1989). Towards an epidemiology of political violence in the third world. Social Science and Medicine, 28, 633-642.

Zwi, R., Radebe, E., Rataemane, S., Freeman, M. & Harris, L. (1995). Mental Health Report – SMT Task Group Gauteng Province. Unpublished policy document submitted to the Strategic Management Team for Health, Gauteng.