The civilising process, violence and race: Thinking with Elias no. 5
Please reference as: Liz Stanley (2017) ‘The civilising process, violence and race: Thinking with Elias no. 5′ http://www.whiteswritingwhiteness.ed.ac.uk/the-civilising-process-violence-and-race and provide the paragraph number as appropriate when quoting.
1.1 This essay is concerned with ‘thinking with Elias’ about matters connected with state legitimacy and the occurrence of violence. It does so in particular in connection with ideas about the importance of the local and contextual in getting to grips with analytical matters, while also not losing sight of the importance of changing ratios power concerning the relationship between established and outsider formations. It is also concerned with the character of events as including sometimes ‘small’ matters of individual lives and deaths, not just so-called a ‘large’ matters connected with the state and long-term processes of change. It does this in connection with murder in South Africa and its changing relationship to race matters.
1.2 The crime statistics for South Africa raise interesting issues including concerning the murder rate, aka intentional homicide rate, and have relevance for the generally shared perception of South Africa as in a number of ways a violent society. They also provide an opportunity for thinking about whether the theoretical ideas of Elias concerning state legitimacy, ratios of power between established and outsider groups and the connections between personality structures and habitus on the one hand and figurations and state formation on the other, might be relevant to comprehending the changing murder rate within the fabric of South African society over time. And relatedly, whether these ideas can provide a helpful means of thinking about its other changes over time is also considered.
1.3 Not surprisingly, there is a good deal of academic work on crime and murder in South Africa. However, discussion here brackets this literature, to think about broad patterns specifically in relation to Elias’s ideas and their utility. There has been substantial South African discussions of these statistics and what follows gratefully acknowledges a number of helpful media overviews that have been drawn on below and can be accessed at the links below. They can be accessed on the Daily Maverick website.
a. Factsheet on South Africa’s 2014/15 murder and robbery crime statistics
b. South Africa’s mysterious murder rate
c. Where murders happen in South Africa
1.4 For some background and a thoughtful discussion of legitimacy issues with regards to the ANC and the South African state, in particular with regard to the amount of violence that occurred around the local elections in August 2016, see Electoral tremors are shaking South Africa’s ANC. How will it respond?
1.5 The discussion that follows was originally written at the end of 2016 and draws on the published crime statistics that has just been published at that point. The most recent national crime statistics were published just before this essay was revised, with the data from this confirming the various trends commented on here.
2. South Africa’s murder rate
2.1 South Africa’s murder rate is high by world standards. The statistics on this released by the South African Police Service (SAPS) in September 2015 put its murder rate at 33 per 100,000 people. This is very high by global standards – the world average rate of ‘intentional homicide’ is around 6 per 100,000. For South Africa, the statistics show that there was a large increase leading up to 1994, a decrease from then until 2012, and an increase over the three years following. The discussion below will consider some factors influencing these changes, then consider aspects of the work of Elias which help in thinking about these matters.
2.2 Different definitions in different countries of the crime involved in the covering term ‘murder’ or ‘intentional homicide’, plus variable data quality, means that international comparisons are not exactly a certain science. But broadly, the highest world murder rates are presently in central and southern America (at the end of 2016, the Honduras rate was about 90 per 100,000, and at the ‘lower’ end that of El Salvador was 41 per 100,000). So, South Africa’s is very high, but not so extremely high. But to put this in further perspective, in 2012 about 3.7% of world deaths due to intentional homicide happened in South Africa, but which had just 0.7% of the world’s population.
2.3 There is a large amount of research and discussion on why South Africa is such a violent society. Explanations largely focus on the ‘usual suspects’, including its history of violence and brutalisation, failures in policing some communities, the impact of apartheid on families, highly unequal and in some places until recently largely non-existent education system, gross material and economic inequalities, structural and individual racism, the wide availability of guns, and so on. But do these help explain the large increase/decrease/smaller increase alluded to above? And what might a take on this informed by the work of Elias add that is useful?
2.4 It was only in 1994 that South Africa’s statistics began to include the circa 16 million people in the old Bantustans (the so-called ‘homelands’). The large increase in homicides in that year shown in the statistics is generally considered to be an underestimate, as policing and proper record-keeping in the ex-Bantustans were often either inadequate or lacking. In addition, in the rest of South Africa much homicide and other crime involving black people as victims or perpetrators had not been reported – and much was not recorded even when it was, as Fatima Meer’s (1976) famous book on Race and Suicide in South Africa pointed out.
2.5 The large increase around 1994 is clearly a data quality matter covering a complex set of politically grounded as well as data issues. And presenting the data as rates per 100,000 doesn’t help in the South African context, because estimates of population size and its distribution have been subject to the same political and data quality issues, and so are also highly unreliable.
2.6 However, analyses which are based on South African Police Service (SAPS) annual reports and figures on crime, and murder specifically, among people living in ‘white’ South Africa from the Central Statistical Service (now Statistics South Africa) and from the Institute for Race Relations show something interesting, albeit with the ex-Bantustans excluded. A 1994 surge does not appear (suggesting that it was indeed the product of changes in the structure of the data regarding the ex-Bantustans), while some longer term trends in the rest of South Africa can be seen.
2.7 First, the part of South Africa under the SAP’s jurisdiction have had murder rates above 20 per 100,000 since at least the 1970s. A high level of violence as shown by murder rates is not a post-apartheid development, but grounded in longer term factors in the South African social order. The need for historically grounded explanations immediately comes into view.
2.8 Second, the same body of data shows that the murder rate surged across the 1980s and 90s to a 1993 peak of circa 78 per 100,000. South Africa was at that time one of the most violent places in the world (and not forgetting that this violence wasn’t equally distributed). Again, there is no certain science here, but certainly a significant proportion of these killings were connected to political conflict, with some of the literature suggesting this was perhaps as high as a quarter; and also, as elsewhere, such a prolonged period of considerable political unrest brought with it increases in the levels of crimes not directly related to politics. Both aspects can be seen as resulting from a decline in state legitimacy, with the first clearly being a direct counter to the white state monopolisation of violence and control, and the second as to a significant extent a by-product of this.
2.9 Third, looking at the data from 1994 on, and particularly from 2003 on when quality issues regarding policing, recording and definitional differences in different areas are much less at issue, shows that the murder rate declined steadily from then until 2012. At that point it had reached a ‘low’ of around 20 per 100,000, which is the 1970s figure for ‘white’ South Africa. This is still higher than in most parts of the world, but half of what it had been. Both the scale and the speed of the decrease are notable. However, while elsewhere decreases in murder rates have also gone hand-in-hand with falling levels of violent crime generally (eg. in New York in the 1990s), in South Africa the picture is patchier. The murder figures correlate with total robbery figures, while there have been significant increases in house and business robberies, although these latter remain comparatively rare.
2.10 Fourth, longer-term socio-political events eventually leading to South Africa’s transition have majorly affected the high murder rate, and then later significantly contributed to the fall in the rate. The decline from 1994 on can certainly be related to the establishment of state legitimacy as perceived by the majority population and the cessation of structural violence and in effect civil war between the apartheid state and the contending political groupings of the ANC and the Inkatha Freedom Party. This was followed by the subsequent consolidation of new state power, and improvements in the social and economic conditions that feed violence which resulted from the new ANC state playing a major role in the redistribution of resources such as housing and education.
2.11 It is worth pausing at this point to consider who it is who is murdered, and who does the murdering. Around 5% of the murders are of children by adults, predominantly males. Around 14% are the murders of women by men, mainly by their present or ex-partners. The remaining 81% are murders of mainly young black men by other mainly young black men. The murders of people who are white by people who are black and the murder of people who are black by people who are white happen, but most are black on black, male on male, young on young.
2.12 So how to explain the increase in the murder rate in each of the years since 2012? The jury is out on this, in the sense that, while it seems that the distribution of the types of murders being committed has changed, why this is so, just what it means and whether the trend will continue, are not clear. Some of the factors involved appear to be as follows.
2.13 Police information – from the dockets filled in by police officers attending crimes – suggests that previously the large majority of murders were committed between people who knew each other and often involved alcohol and/or drugs – so-called social murders. But by the mid to late 2000s, these as a proportion had declined and what had risen were murders committed by strangers, often in the commission of other crimes such as robbery and burglary – so-called instrumental murders.
2.14 Also, there has been an increase in politically-motivated murders and also those occurring in the wake of political unrest and violence, which appear to be directly related to a decline in state legitimacy and an increase in clashes between members of contending power groups. Relatedly, the so-called instrumental murders may also be indirectly related here, with the new state being now seen as ineffective in distribution and culpable regarding widespread structured corruption, leading to a loss of confidence that there will be further distribution and generational improvements, including with regards to employment and income.
3. Thinking about this with Elias
3.1 The patterns that exist in South Africa’s murder rates and briefly sketched above call out in Eliasian terms for a detailed analysis of historical trends and the ‘local’ circumstances – of South Africa, of times and places in South Africa, of different peoples and communities in South Africa, of particular events and occurrences too – which have given rise to these patterns. Among other things, this is to emphasise the importance of the contextual and situational, because the changes in these patterns briefly noted above each have different influences and outcomes. While Elias is always attentive to the broad trajectory of developmental patterns over time, he is also alert to the importance of and the differences made by the local and situational, and that apparently unidirectional outcomes can actually contain diverse and even contradictory movements.
3.2 Elias is also clear that ‘the individual’ does not exist as a sole isolated entity (the homo clausus), but is by definition or constitution part of a ‘we’ and best understood in terms of an ‘I-and-we’ ontology. The patterned character of the murder rate, in South Africa as elsewhere, demonstrates just as clearly as the patterned character of the suicide rate does that killing like other social behaviour is precisely social and related to prevailing normative ideas and beliefs, social and economic circumstances, and political frameworks. What this suggests is that any Elias-inspired historical investigation of the long-term incidence and meaning of murder (and other violence) should proceed by looking at paradigmatic models or examples over time and in particular places and circumstances, rather than homogenising these in generalised terms or producing only snapshots at particular time-points.
3.3 People’s behaviour occurs in contexts which are both local and supra-local, and for Elias there is a fundamentally important relationship between social habitus and developments at the supra-local level of state formation and activity. Long-term patterns of state formation and dissolution around the establishment or perpetuation – or rejection – of state legitimacy, the state monopoly of (legitimate) force, and its successes or failures with regard to the accumulation and distribution of resources, are on the agenda in his investigations of how civilising processes pan out over time. This is in particular through the relationship of these things to habitus and the existence and organisation of groups within key figurations in the social order.
3.4 With respect to the South African context and its murder rate, it is clear that many intentional homicides over the decades have resulted from political factors and from circumstances brought about or significantly and negatively impacted by political factors. They are directly related to matters of state formation, challenges to the state, how the state endeavours to legitimises itself, and how it handles matters of accumulation and distribution of resources.
3.5 This is of course not a matter of the state having impact by its activities somehow determining habitus and personality structures and so shaping behavioural predispositions, a deterministic kind of approach that Elias completely eschewed. For instance, differently composed groups and their figurational relations will respond differently to the same social circumstances, different local contexts and their histories will also play their part, and so on. However, at the same time the supra-local framework of economic, political and also symbolic possibilities, including concerning resource distribution, is important and needs to be taken into full account.
3.6 For Elias, a key matter to be considered in any social investigation is the interrelationship between an established and outsider groups in the networks persisting over time that he termed figurations (and see here the ‘Thinking with Elias essay on De/civilising processes). Over the period covered by the statistics discussed earlier, at the level of the state there has been a reversal of the old established group (the National Party minority white state) and its key outsider group (the ANC-headed majority black state). And in the wake of this, there has been the creation of new established-outsider relations, a new figuration in which formations based around whiteness are increasingly infrequent and mainly occur in some economic contexts and in the organisation of private life.
3.7 What appears to be happening now, over the last few years in which the murder rate has increased, is a kind of radical disenchantment, a loss of political faith, on the part of many people concerning the possibility that the ideals of the pre-1994 period would be fully enacted after the political transition had occurred. Perhaps more to the point, this has been accompanied by major political disturbances, including within-group conflicts, as well as more ordinary forms of protest, and some of these have been accompanied by violence and loss of life.
3.8 The state, then, is not removed and operating in an outside context. Instead in direct and indirect ways it connects with the character of the developing habitus at local levels, and the beliefs and normative standards that are a part of the personality structures associated with the groups composing figurations. The disenchantment referred to above is one in which the ANC state is now simply the state, and a state associated with corruption as well as a loss of faith concerning economic and political transformation. And at many local levels, the conditions which in earlier pre-transition periods gave rise to hopelessness and violence are to a significant extent still present.
3.9 Events, even when seemingly individualised ones such as murder, are shaped by social structures and embedded in unfolding social processes. Understanding them and their meaning requires attention to both the structures and the proceses and indeed to see these as intertwined. Elias provides some helpful tools for doing this,
- in rejecting the homo clausus idea and centring the I-and-we relationship,
- in relating personality structures to the formation and re-formation of habitus,
- in focusing on the centrality of groups within figurations,
- in highlighting power ratios in the relationship between established and outsider groups and treating these as composing one figuration,
- in rejecting cross-sectional snapshots and favouring historically grounded and over time investigations,
- in insisting on the importance of the grounded and specific while also taking into account the supra-local, including the import of the state and its role in accumulation, monopolisation and distribution as well as regulation,
- and in asserting the impact of such things on all aspects of the social order.
4. Ordinary murderings
4.1 Briefly, a number of murders or incidences of murder will be considered within an Eliasian framework informed by the above points. The intention is to think broadly about the utility of the framework.
Where and who: murder rates
4.2 50% of South Africa’s murders in 2014/15 occurred in 12.3% of police precincts. Some 20% of police precincts in more affluent metro and rural areas and towns had a murder rate of less than 12 per 100,000. And just over 10% of policing precincts had a murder rate of zero.
4.3 One of the facts arising from the 2015 release of statistics (with little to no change having occurred in the 2016 statistics) is that Cape Town murder rate is considerably higher than that of Johannesburg and Pretoria. Indeed, police have recorded more murders in Cape Town than in Johannesburg and Pretoria combined. In effect, Cape Town residents are almost twice more likely to be murdered than Johannesburg ones. But this begs the question of what is meant by ‘Cape Town’ and also who its residents are in race, gender, age and occupational status terms.
4.4 Over 60% of Cape Town murders occur in ten of its sixty police station precincts, with six of them – the Nyanga, Harare, Mitchellsplain, Gugulethu, Khayelitsha and Delft precincts – having had persistently high murder rates for more than a decade. This is over the same period of time that the murder rate overall declined, and then over the last three years has increased. This suggests that local factors are at work, perhaps influenced by supra-local ones of the kind discussed earlier. In particular it indicates local circumstances, embedded in relationships between people and groups, and in different areas of these precincts and townships, of a kind that Elias discusses in The Established and the Outsiders.
4.5 Commentators on this have focused on gang cultures and a cycle of deprivation and violence (eg. Don Pinnock Gang Town (2016) Tafelberg Publishers). An Eliasian take on such things might focus instead on the more ordinary functioning and activities of established and outsider groups, the figurational relationships involved and how these have changed over time. Relatedly, there are degrees and kinds of ‘outsiderness’; and in some areas people who are the ‘majority of the ordinary’ have developed organisational structures to enable them to exert greater protection from if not always control over locally contending established-outsider groupings.
Death of an ordinary member
4.6 What has been called in the mass media ‘the death of an ordinary member’ occurred in mid June 2016. A local election campaign was in the making, concerned with the metropolitan area of Tshwane and who would be its mayor for the forthcoming period. The context was one of ANC majority control of the metro and the policy of candidate lists, with voters choosing a party rather than specific candidates.
4.7 The ANC candidate list proved contentious, because the sitting mayor was not named and nor was his chief rival. Listed instead was an ‘outsider’, a woman who in fact had lived in the area for over 20 years but who came originally from Natal. Violence erupted in one of the townships around Pretoria over several days, largely between supporters of the two men although with ethnic aspects too, and was in effect a factional small war. It involved a series of demonstrations and other events including crowd rampages, the burning of vehicles and buildings – and also gunshots exchanged between groups contending over particular candidates omitted from the list. At one of these occasions outside the Tshwane Events Centre, shots were fired and several men were injured, with one ‘passing away’, the euphemistic term used for internecine murders. As the availability of such a descriptive term indicates, this particular death did not occur in isolation, for other internecine murders had occurred previously at orchestrated political events in which violence took place between politically contending groupings, sometimes being confined to them but sometimes lapping over onto bystanders.
4.8 Certainly established-outsider relations were involved in this Tshwane death, but in a quite complicated way. These events have involved factions within the established ANC group, who orchestrated political events which resulted in (and may have been intended to result in) not only the symbolic violence of burning vehicles and buildings but also the person-to-person violence of guns and shooting. And because of the relationship between the ANC and the state, both locally and supra-locally, these events are connected in a direct way with the role of the state with regard to matters of legitimacy and also the accumulation and distribution of resources, not to mention departures from the regulated way in which the state might be expected to control and contain violence.
4.9 The rather bland phrase ‘changing ratios of power’ in this instance stands for some very public eruptions of symbolic and physical violence, and an investigation of established-outsider relations within an Eliasian framework can proceed from such events, to investigate the structures and processes which give rise to them.
See ‘The murder of an Ordinary Member‘, June 2016.
Two farm murders
4.10 The farm murders in South Africa which often catch media attention and are assumed to be prototypical are of white farmers attacked by their workers or by marauding groups. There are around twenty or so of these incidents a year and most involve murder or murders committed in the course of robbery. Less well reported are the murders of black farmworkers by white farmers, of which reports there are fewer, not necessarily because fewer such deaths occur but because they are easier to cover up under the heading of ‘disappeared’. In a case when WWW research began in South Africa, for instance, a group of white farmers hunted for ‘fun’ and then shot a farmworker and left his body to be eaten by wild animals.
4.11 In Parys in mid June 2016, at approximately the same time as the Tshwane murder discussed above, four local white farmers and a white police captain were arrested and bailed concerning the murder of two black farmworkers. They had earlier concocted the story that the farmworkers had attacked one of them. It seems instead that the men had called to request outstanding payment for work done by one of their sons, the farmers responded with fury perhaps fuelled by alcohol and attacked them, and attempts to cover this up were aided by the police officer. Demonstrations around the courthouse when the first hearing was held to charge the four farmers involved a stand-off between a group of local white farmers singing the old apartheid national anthem and waving ancient Transvaal and Free State flags, and ANC and Economic Freedom Party demonstrators, with bottles and other missiles thrown on both sides.
4.12 If violence is engrained in ordinary ways of behaving and in this sense is endemic in a society, then it is also endemic in the sense that it affects everyone, because people are interlinked with each other in figurational relationships. In the case of South Africa, one notable indication here concerns driving habits and a ‘drive to kill’ approach particularly on major roads, something that makes the regulated politeness and turn-taking of its four-way stops atypical rather than indicative. South Africa’s murder rate is high, and so too are its deaths on the road and the fearsome toll taken of human life on every public holiday.
4.13 One such engrained relation concerns the vast disparities of apartheid-era relations between white and black people. This was perhaps particularly in farming, mining, domestic labour and other contexts in which there were inbuilt stark inequalities between the established white group and black people treated as outsiders around contemptuous ways of behaving towards them but which was seen as justified by their ‘inferiority’. It is in this context that the events in Parys need to be placed, while keeping firmly in mind Elias’s analysis of former established groups as inhabiting a kind of collective fantasy about their superiority, and a timelag of possibly a number of generations in its members coming to terms with the loss of position and fall in the social hierarchies now prevailing. The turnout at the commital hearing of the Parys farmers accused of murder, involving local whites singing racially-loaded anthems and waving flags belonging to the old Boer Republics, is a clear reminder of the relevance of Elias’s analysis of this.
4.14 In many rural areas, ‘old’, pre-transition, established white and outsider black groups remain still locked into unequal relationships in their local dealings with each other. This has been perpetuated because of the continuation of circumstances in which some whites employ black people around economic relations premised on binary notions of superiority and inferiority. ‘Ratios of power’ is no simple matter and there can be differences between those power relations prevailing in one area of people’s lives, and those prevailing elsewhere. Elias’s comments about the ‘collective fantasy’ of former established groups and the related sometimes generational timelag in becoming accustomed, if not reconciled, to a new ranking are highly pertinent here.
5. What does thinking with Elias add, or bracket?
5.1 Starting with what is bracketed, there is what for many readers may be the uncomfortable fact that race as a ‘thing in itself’ producing inequalities, discrimination and contemptuous views of the ‘other’ is seen by Elias as a rationalisation, covering power disperities used to extend and maintain group controls over resources. This is the same approach that Elias has regarding the underpinnings of other structural inequalities including those of sex, religion, ethnicity and class. However, for some people who otherwise advance broad constructionist ideas, on one level recognising the socially constructed character of race can be combined with on another assigning it a privileged status. For Elias, the analytical focus should instead the on the underpinning dynamics of how disparities in power relations pan out over time and impact on established and outsider relations. This ensures that his analysis is as capable of investigating within-group disparities as those between groups. It also means there are no covert essentialisms or claims to epistemological privilege lurking within the framework of his ideas.
5.2 While Elias recognises the individual prejudice can exist (in which person A is prejudiced about the specific characteristics of person B), analytically he is not concerned with this, but instead with the established-outsider relation and how groups within this are seen to be a priori associated with a set of definitional either ‘superior’ or ‘inferior’ characteristics. This is a matter of social and power relations and disparities around the membership of interrelated but conflicting groups. Thinking in this way ensures that both the social basis of structural inequalities, and the possibility of long-term changes in these patterns relationships within a figuration, are foregrounded rather than treating such things as psychologically-based and so resistant to change.
5.3 Neither social structures nor the operations of the state are treated in a removed or abstract way within the framework of Elias’s thinking. State formation processes, challenges to the state, and the role of the state in the accumulation and distribution of resources and the monopolisation of force are not ‘elsewhere’. They are part of the social fabric and interface in many ways, times and places with the lives and behaviours of social groups and individual people. This includes established-outsider relations, and indeed in some circumstances these can be formed around emanations of the state. For Elias, then, it is not just that the state ‘influences’ or ‘impacts’, but that it is a constituent part of the unfolding character of events and circumstances and its presence can take a number of different forms.
5.4 Rejecting a ‘snapshot’ cross-sectional approach to social investigation, Elias’s concern is with the long-term and sociogenesis, the processes of social becoming and changes over time. For him, ‘snapshot’ approaches encourage seeing people in homo clausus and over-individualised terms, and also inhibit full recognition of the in media res character of social life. Things change over time; nothing remains exactly the same because if nothing else the context and the persons involved change. Put rather baldly, if temporality is not part of a research design then the design is not doing its sociological job properly.
5.5 The idea of a figuration is a helpful and a powerful thinking tool, even while recognising operational issues in pinning down exactly where the boundaries of a figuration lie and the cut-off points in terms of membership. Elias’s approach is to work at sufficiently small-scale as to be able to isolate boundaries and memberships, in the context of the discussion here concerning the established and outsiders figuration. In particular, the emphasis on groups locked into an unfolding relationship around dynamics set in motion by power disparities and the control over resources is particularly helpful. So too is his twin recognition that there will be changes in power disparities over time and that for a former established group this will involve sometimes considerable timelags in fully grasping their changed position within social hierarchies.
5.6 The discussion here has taken the example of murder in South Africa and used this as a pivot point for some thinking about Elias’s famework of ideas and their utility or otherwise, in particular regarding the established-outsider relation, figurations, and ratios of power. Elias’s work provides useful ideas for thinking through events and the social structures and social processes these are embedded in. Depending on a particular research investigation, it may need additions or perhaps subtractions. However, this is a point that Elias himself makes, that times, places and persons all make a difference, and the specificities of particular circumstances and situations are important and need to be fully taken into account.
5.7 As this last comment indicates, although Elias’s ideas were conceived around sociogenesis and the civilising processes of Europe and specifically France, Germany and Britain, they can be turned for use in thinking about South Africa in ways that highlight rather than diminish its particular circumstances and specificities.
Last updated: 13 May 2021