The Pathway to Murder:
A Social Psychological Study of the
Evolution of Violence in an Industrial Disputeby Lloyd Vogelman
Chapter Two: Literature Review and Model
Introduction
Collective violence assumes a multitude of forms: group ritual murders, gang rape and military killings are some of the most common examples. This diversity has stimulated academic specialisation in the various manifestations, and in part this is an acknowledgement that despite similarities, each form of violence has distinctive causes, power dynamics and effects. This study is primarily concerned with the reasons for the emergence of collective violence within an industrial dispute, and the factors that encourage certain individuals to participate in such behaviour.
Virtually all of the disciplines which fall under the broad rubric of "social science" explore the subject of collective violence and the academic literature concerning this subject is both substantial and comprehensive. Despite this extensive academic analysis, however, there is still no single theory that thoroughly explains the phenomenon or its various manifestations. For every explanatory law constructed about the aetiology or processes of collective violence, there is an exception.
Since I have a dual focus in this study, namely to comprehend the aetiology of the Eight's violence from both an individual and a group perspective, I have drawn on a wide range of literature. Introductory chapters to general books on violence frequently explore the aetiology of violence according to biological, psychological, sociological and socio-psychological theories (McKendrick & Hoffmann, 1990; Naude & Stevens, 1988). However, these texts do not adequately acknowledge that with the probable exception of some biological theories, the majority of theories on violence are not mutually exclusive. Because of the interconnections between them, categorisation of the different theories on violence is a formidable endeavour. However, categorisation is useful because it provides a framework within which to decipher the abundant expositions on violence and the major distinctions between them.
Some of the categories that I have proposed, while they still fall within the broad ambit of the biological, psychological or sociological, are more specific and may include literature from a range of disciplines. Thus in the "developmental" category, both psychological and sociological theories are included. In reading the literature review that follows, it should be borne in mind that space constraints permit me to mention only a few of the numerous perspectives on the causes of violence. Although this literature review is mainly concerned with the literature on violence, there are other literatures that are also pertinent to this study, for example, those dealing with poverty and strikes. These will be explored in later chapters of the study.
It should be noted that some of the literature that is reviewed does not make a clear distinction between aggression and violence (Fromm, 1974; Storr, 1970). There is, however, an important distinction between the two: aggression, unlike violence, does not involve the use of physical force (Siann, 1985). The importance of this distinction will become more apparent later in this chapter when I outline my model of violence.
1. Literature Review
1.1 Theories of Aggression and Violence
1.1.1. Biological Theories
The type of violence that took place at Prolecon is often described as "inhuman", as it was brutal and seemingly without compassion. Presumably, many of those who use the epithet "inhuman", define themselves as human, and so this epithet may serve to reassure them that they are incapable of such violent conduct (Siann, 1985). This label also serves to enhance the perception that perpetrators such as the Eight are biologically distinct, and thus it contributes to the avoidance of any consideration of the psychological and sociological factors which may have contributed to their conduct.
Lombroso (1876) conducted one of the first systematic studies on the biology of criminals. He asserted that the traits of subordinate animals and primitive men re-emerged intermittently in certain individuals. These traits were exhibited in degenerative bodily characteristics such as ears without lobes and receding chins. Although, as Siann (1985) points out, few present day theorists are likely to endorse such simplistic and crude arguments, biological theories of violence continue to be popular. Some of these theories, and particularly those which explain violence in terms of brain pathology, neurotransmitters, hormones and human instinct will be briefly considered below.
As I have no expertise in the field of biology or medicine, and because this thesis is concerned with social psychological factors that contribute to violence, my examination of some of the biological theories of violence is extremely cursory. I have, however, chosen to address this literature because it comprises a fundamental aspect of the literature on violence, and without this aspect, the literature review would be incomplete. The first group of theories that will be I discussed are those related to brain pathology.
1.1.1.1 Brain Pathology
At the most rudimentary level, these approaches argue that violence is shaped by the physiology of the brain (Bylinsky, 1982; Moyer, 1976; Siann, 1985). For example, in animal (Egger & Flynn, 1963; Johnson, 1972) and clinical (Shah & Roth, 1974) studies, researchers have cited relationships between lesions in the limbic system and violent behaviour.1 Mark and Ervine (1970), for instance, found that 75 percent of their sample of 400 violent prisoners had suffered serious head injuries, and consequently they postulated an association between the two variables.
1.1.1.2 Neurotransmitters
Neurotransmitters, and in particular noradrenaline, dopamine and serotonin, are believed to transmit messages "associated with motivated behaviour either in the bloodstream or in the nervous system" (Siann, 1985, p. 32). All three of these neurotransmitters have been linked to violent behaviour in animal studies (Moyer, 1976; Vowles, 1970). In human studies (Bioulac, Benezech, Renaud, Noel and Roche, 1980; Bylinsky, 1982), research suggests that some neurotransmitters do evidence a more consistent positive correlation with violence than others. For example, there appears to be a general consensus that noradrenaline has a stronger relationship to violent behaviour than serotonin (Shah & Roth, 1974).
1.1.1.3 Hormonal Theories
Amongst the most popular of the biological theories are those that relate hormones to violence. More particularly, research in this area has attempted to provide an explanation as to why men are more violent than women (Archer, 1976; Lunde & Hamburg, 1974). In essence, these theories contend that androgen, the dominant hormone in males after the age of ten, is more closely associated with aggression than is estrogen, the hormone which becomes dominant in females at the same age (Archer, 1976; Lunde & Hamburg, 1974).
1.1.1.4 Assessment of Biological Theories
Earlier in this chapter, I indicated that I do not have a background in either biology or medicine, and that I have not studied biological theories in any depth, and as a result it is difficult for me to assess the validity of these theories. Siann (1985), however, in her description of these theories, provides some useful insights about their limitations. First, some of the research findings on which these theories are based are questionable, because of problems relating to methodology and sample size, among others. Second, humans can not be reduced solely to physiology or chemistry. There are numerous other psychological and sociological factors that influence behaviour, and it is not always possible to separate out the biological from the psycho-social factors. For example, it is almost impossible to conclude whether the violence of the sample of prisoners in Mark and Ervine's (1972) study was caused by head injuries, or by the material hardship that is likely to have engulfed their lives. Third, while there seems to be little doubt that certain brain pathologies are a major cause of violence in some individuals, not all of those individuals with brain pathology are violent. Consequently, it is necessary to investigate to what extent situational, family history, and social learning factors inter-relate with biological pathology to produce violence.
Within the domain of collective violence, biological theories would appear to have limited applicability. It seems absurd to postulate that the political violence committed by black South Africans during the apartheid years was the result of brain dysfunction or hormones, as opposed to political, economic and social dissatisfaction with apartheid. Similarly, in this study, if brain lesions were responsible for the violence of the Eight and other strikers, it would be necessary to explain how it was that all those with brain lesions converged at the same time, were angry about similar things and expressed their violence towards similar targets? These points may seem obvious, but they need to be articulated, as there are biologists who have asserted a possible connection between brain dysfunction and collective violence. Mark, Sweet and Ervine (1967) argue, for instance, that the urban riots which plagued the United States in the 1960s may have been the result of undiagnosed brain pathology. To reduce collective violence, they proposed clinical investigations in order to identify and treat individuals with low violence thresholds.
Some crowd theorists (Tarde 1893; Sighele, 1895) have argued that crowd violence is frequently linked to the inherent moral pathology of crowd members. This pathology was often explained in terms of gender. For example, Tarde (1893) and Sighele (1895) believed that a crowd composed of women was more malevolent than its male counterpart. This argument was, as Barrows (1981) points out, used to justify the continued "civil and domestic subjugation of women" (p. 149) in Europe at that time.
Although biological theories of violence cannot explain collective violence, they may, albeit minimally, assist in explaining the different behaviours within a crowd, including why some individuals commit violence - or commit more violence - than others. In this study, where not all of the Eight were equally violent, there is the possibility that physiological factors did have some influence.
Biology may also contribute to violence more indirectly, through possible inherited attributes such as temperament and activity levels. These attributes play a part in shaping not only an individual's behaviour, but also the responses that he or she receives from others (American Psychological Association [APA], 1993; Earls, 1991). For example, a difficult temperament may negatively affect the emotional bond that an infant establishes with his or her parents (Earls, 1991). This in turn may increase the potential for emotional neglect or physical abuse - childhood experiences, as I will indicate later in this chapter, which are associated in the psychological literature with violence. While these and other biological issues are not the focus of this study, it is necessary to acknowledge that they may have played a role in the evolution of the Eight's violence. There is one other dimension of biological theory that I would like to report on before leaving the category of biological causation - the ethological.
1.1.1.5 Ethological Approach
The ethological approach is founded on the study of animals within their natural habitat, rather than in a laboratory setting, and it draws on Darwin's evolutionary theory rather than physiological pathology (Siann, 1985). This approach is best characterised by the work of Lorenz (1969), Morris (1967) and Ardrey (1972). Generally, these theorists propound that human behaviours, including violence, are instinctive. In addition, they contend that human nature is constant and immutable. Morris (1967) writes:
Behind the facade of modern city life there is the same old naked ape. Only the names have been changed: for "hunting" read "working", for "hunting ground" read "place of business", for "home base" read "house", for "paid bond" read "marriage", for "mate" read "wife" and so on. (p. 43)Lorenz (1969) is the most influential of the ethologists, and it is to his work that I will turn my attention. Essentially, Lorenz (1969) posits that aggression is an inherent predisposition rather than a product of social experience. He argues that aggression, because it is an intrinsic predisposition like other instincts such as hunger, sexuality and flight, has its own energy which must be discharged. Thus aggression is not a response to external stimuli, rather it occurs because its energy must be released. These and other findings, such as those that animals control their aggression through appeasement gestures, have been severely criticised on empirical grounds (Barnett, 1973; Schuster, 1978; Swanson, 1976). Lorenz, Siann (1985) writes, has not only "misrepresented some of the ethological findings" but also "nearly all of the anthropological findings" (Siann, 1985, p. 65) on which his conclusions are based. However, despite the problems with which Lorenz's findings are associated, it would seem that because humans do commit violence with such frequency, they do have a capacity for the behaviour. However, social conditions are likely to influence how that capacity is exercised, and whether it is exercised at all (Klineberg, 1981).2
1.1.2 Developmental Theories
These theories stem largely from the psychological literature, and they tend to focus on emotional, cognitive and behavioural development in the early years of life, and the ways in which this development leads to violence and particular characteristics (such as hypersensitivity to criticism) amongst violent offenders. The psychoanalytic literature, drawing on the work of Freud (1932), places strong emphasis on violence as a product of the innate drive of aggression (Fromm, 1974; Storr, 1970). However, unlike Lorenz (1969), Freud (1932) and later psychoanalysts such as Storr (1970) suggest that the way in which this drive manifests is dependent on the individual's early emotional history and the unconscious motivations that this history produces. For example, Freud (1909) argued that if an infant's oral desires were frustrated, the infant was likely to become aggressive and that later on in adult life this aggression may be exhibited in a tendency towards "biting criticism" (Kutash, 1978, p. 9). Freud's assertion that frustration precipitates aggression formed the basis of one the most notable theories on aggression: Dollard et al.' s (1939) frustration-aggression hypothesis. This hypothesis is central to the model of violence that I have constructed, and will be discussed in section 2.
Adler (1927), a colleague of Freud, posited that feelings of helplessness that originated in childhood precipitated an inferiority complex, and encouraged the child to compensate for these feelings through striving for power and control. Storr's (1970) work Human Aggression combines both an ethological and a psychoanalytic perspective, and in a similar vein, he argues that the resentment engendered by a sense of impotence in childhood is likely to be converted into malice and enmity in adulthood.
The view that childhood experiences influence later adult behaviour, and more particularly violent behaviour, has become increasingly accepted within the psychological literature. Miller (1987) traces the roots of violence to child rearing practices, and she highlights experiences such as physical abuse, a denial of the child's rights to express him or herself, and a lack of education. She writes that "every act of cruelty, no matter how brutal and shocking, has traceable antecedents in the perpetrator's past" (p. ix). To substantiate her thesis, she provides a case study of Hitler's life and describes the abuse that he endured as a child. Miller's work is problematic, in that it fails to account for why it is that some individuals with histories of neglect, repression and torment do not engage in extreme violence as adults. Miller does not acknowledge the different means by which children cope with childhood trauma, based on differences in their constitution, familial dynamics and support networks (Dawes, 1987; Straker, 1992). She also fails to locate Hitler within a particular political, economic and social context that promoted and sanctioned his racist and authoritarian doctrines. Fromm (1974), a psychoanalyst with a strong social orientation, addresses this aspect particularly well in his case studies of Hitler and Himmler.
In addition to Miller and Fromm, a host of other social scientists view violence as a principal means of compensating for emotional deficiencies during the childhood years (Adler, 1927; Gilligan, 1991; McKendrick & Hoffmann, 1990; Pizzey, 1974; Renvoize, 1979; Toch, 1972; Walters, 1975). Empirical investigations would appear to confirm this assertion. For example, Frazier (cited in Gilligan, 1991) demonstrated that violent men often experience a high level of humiliation in their early years. Toch (1972) in his study, Violent Men: An Inquiry Into the Psychology of Violence, argues that many men who are prone to violence have experienced interpersonal relationships in their early years in which there is a deficiency "in stability and emotional support, thus making it difficult for positive self perceptions to develop" (p. 231). For these men, violence is a means of fulfilling their needs for power, control, uniqueness and importance.
One of the more polemical debates in psychological inquiry concerns the effects of a violent socialisation on adult behaviour. In line with popular opinion, much of the literature - from both psychological and sociological sources - asserts that the abused become the abusers of the next generation (Guttmacher, 1967; Hoffmann & McKendrick, 1990; Pizzey, 1974; Renvoize, 1979; Walters; 1975). This is a recurrent theme in studies concerning delinquency (McCord, McCord, & Howard, 1961; West & Farrington, 1973, 1977); wife battery (Pizzey, 1974; Renvoize, 1979), and physical assault (Toch 1972).
There is also much debate concerning the process through which the abused may become the abusers. Learning theorists assert that individuals learn to be violent through observation, modelling and reinforcement of violence in their childhood years (Bach-y-Rita & Veno 1974; Bandura 1973; Bandura, Ross & Ross, 1963; Buss, 1961, 1963, 1966; Buss & Brock, 1963). Other psychologists have used the psychoanalytic concept of identification with the aggressor to explain the acquisition of violent behaviour (Miller, 1987; Vogelman 1990a; West, 1985). This theory suggests that individuals enact the same violent behaviour as their abusers because this behaviour enables them to eschew the feelings of powerlessness that they experienced as children. Both of these theories will be explored in chapter six. In general, however, there is widespread agreement that socialisation patterns in which violence is the primary means of discipline foster the conviction that violence is an appropriate method of resolving conflict and achieving goals in other domains. Behaviour that is learnt in micro situations may be used in the same micro location, or may be generalised to other macro principalities of life when problems or conflicts arise.
Although they agree that childhood history is a fundamental determinant of adult behaviour, Widom (1989), Straker (1992), Dawes (1987) and a host of other researchers contend that the straightforward proposition that childhood experiences are imitated in adulthood is too rudimentary. These writers base their argument on the fact that numerous children exposed to violence in their early years do not exhibit violent behaviour at a later stage. Straker (1992), in her book Faces in the Revolution, cites case examples of black youths in the township of Leandra who, despite their violent socialisation, exhibited high levels of compassion and social consideration as adults. While the debate concerning the exact effects of a violent socialisation will probably never be conclusive, it is safe to assume that a violent socialisation does increase the potential for later use of violence, especially if the individual is located within a culture and subculture that endorse violence. This point brings me to the third theoretical category: cultural theories.
1.1.3 Cultural Theories
1.1.3.1 Cultures of Violence
This approach proposes that violence occurs partly because there is a general cultural acceptance of the behaviour (APA, 1993; Australian Institute of Criminology, 1989; Hoffmann & Mckendrick, 1990; Klineberg, 1981; Wilson, 1983). This acceptance is reflected in the different ways in which violence is woven into the cultural fabric of the society through its presence, for example, in film, television (APA, 1993; Klineberg, 1981) and sport (Siff, 1990); and the social sanction given to violence through corporal punishment (Holdstock, 1990), the death penalty (Simpson & Vogelman, 1989) and access to firearms (APA, 1993). This view suggests that violence is not only condoned and prized, but also learnt through this social sanction, and so it is strongly associated with the literature focusing on socialisation (Pizzey, 1974; Renvoize, 1979), modeling and social learning (Bach-y-Rita & Veno 1974; Bandura 1973; Bandura, Ross & Ross, 1963; Buss, 1961, 1963, 1966; Buss & Brock, 1963).
1.1.3.2 Subcultures of Violence
Wolfgang and Ferracuti's (1967a) theory of a subculture of violence is one of the most renowned expositions on the social acceptability of violence. This theory attempts to identify characteristics of those communities and societies which are more beset by violence than others (see chapter six). In so doing, it draws on two main sources: Durkheim's (1951) Suicide and Sellin's (1938) Culture, Conflict and Crime. The former, a study of anomie, is concerned with the degree of the individual's integration into social groups, while the latter work explores the operation of norms.
Wolfgang and Ferracuti assert that in order to understand subcultures of violence it is necessary to examine their value systems; the importance of life in the scale of their values; expected reactions to different stimuli; perceptual differences in evaluating stimuli; and the general personality structure of the individuals within the subculture. A principal flaw in this theory is that it does not examine the origins of subcultures of violence. Wolfgang and Ferracuti (1967a) merely remark that there may be different origins in different cultural settings, and conclude: "We are not prepared to assert how a subculture of violence arises II (p. 260). Consequently, their theory is ahistorical and devoid of sufficient emphasis on some of the social, economic and political elements which underpin any subculture of violence.
Part of the value of Wolfgang and Ferracuti's theory of a subculture of violence is that it is able to explain why violence is maintained once it emerges. The theory touches on the issue of group dynamics and the ways in which these dynamics help to conserve a uniformity of behaviour, but this theme is not developed. I will I return to the issue of group dynamics in section 1.1.4, in my discussion of the literature pertaining to crowds. First, however, I wish to consider another aspect of the literature that is related to a subculture of violence: the psychological and social rewards of committing violence within a culture, subculture or group which condones the behaviour.
1.1.3.3 The Psvchological and Social Rewards of Committina Violence
The perpetration of violence within the context of a collective that endorses the behaviour has a number of potential advantages for the individual's position within the group. In this situation, violent behaviour embodies two ingredients vital to social identity: it confirms the individual's commitment to the group's norms and validates his membership of the group (Tajfel, 1981). Violence also helps to facilitate bonding within the group. These benefits are adeptly characterised in Groth's (1979) analysis of gang rape:
One of the unique dynamics in gang rape is the experience of rapport, fellowship and co-operation with the co-defenders It appears [as if the offender] is using the victim as a vehicle for interacting with other men … behaving … in accordance with what he feels is expected of him … validating himself and participating in a group activity. (p. 113-115)Arendt (1963), in her thesis on the banality of Nazi evil, argued that in certain contexts violence becomes a means of social advancement. The contention that many of the Nazi atrocities were motivated not merely by sadism but also by the banal desires for occupational, social and political promotion, provoked heated debate and acute criticism from authors such as Bettelheim (1963), who were prominent in analysing and commenting on the Nazi genocide of the Jews. According to Fromm (1974), in his analysis of Nazi atrocities, the emergence of mass scale sadism and brutality is also linked to the existence of sadistic personalities who have the opportunity to thrive because their societies condone brutality towards certain groups. Fromm (1974) in his case study of Nazi leader Heinrich Himmler, writes:
A person whose character is sadistic will be essentially harmless in an antisadistic society; he will be considered to be suffering from an illness. He will never be popular and will have little, if any, access to positions in which he can have any social influence. (p. 298-99)Later in the case study he advances this line of argument:
There are thousands of Himmlers living among us. Socially speaking, they do only minor harm in normal life, although one must not underestimate the number of people they damage and make thoroughly unhappy. But when forces of destruction and hate threaten to engulf the whole body politic, such people become extremely dangerous; they are the ones who yearn to serve the government as its agents for terror, for torture and killing We must come to the conclusion that the Reichsfuhrer SS was a sadistic character before he was a Reichsfuhrer; his position gave him the power to act out his sadism on the historical stage. (p. 323)For Fanon (1968), the commission of violence by the oppressed against their oppressors is not only an act of outrage and valour but also one which communicates a message of hope to others, similarly oppressed, that exploitative social relationships can be challenged and overturned. In this respect, violence takes on a more universal social importance. War is the most brazen example of a context where mass killing is viewed as socially constructive, as soldiers on both sides believe that killing will assist in safeguarding sacred values and interests. Lifton's (1986) work on Nazi medical experiments highlights another aspect of the positive perception of mass killing. He argues that the doctors who committed medical abominations on children, the mentally ill, the mentally handicapped, the physically disabled and the racially inferior believed that they were making a meaningful contribution to medicine and the advancement of the German volk.
A principal ethos of war, political struggle or any other conflict is that courage and violence committed in the name of a cause is considered tenacious and brave. These attributes are also strongly related to another benefit accrued through violence: the affirmation of male sexuality. Violence and masculinity are characterised by many similar features: control, power, and physicality (Campbell 1991; Hoch 1979; Glaser, 1986; Toch, 1976; Vogelman, 1990a; Wolfgang & Ferracuti, 1967a). Although affirmation of masculinity is important for men of all classes and social positions, it is especially significant for men who are oppressed, as they are denied sources of esteem that are available to the privileged (Toch, 1976). These men tend to rely more strongly on their masculinity to acquire status and power (Toch, 1976). The quest to affirm masculinity may be manifest in sexual promiscuity (Vogelman, 1990a), employment traditionally associated with maleness (Tolson, 1977), and the commission of violence. Violence is also an act which protects men from allegations of weakness or femininity - allegations which can be particularly damaging to self esteem. For those men already victimised by such "slurs", violence offers a hasty return to "remasculinisation", social acceptance and self respect.
The use of violence as a means of reaffirming masculinity and enhancing physical vitality is not only the province of vampire tales. Numerous forms of violence, from the criminal to the political, utilise violence in this way. In South Africa, for instance, there have been incidents where perpetrators of political violence have imbibed the blood of their victims so as to become revitalized (Weekly Mail, 5-11 June 1992). This revitalisation is perceived as essential in order to maintain strength for future battles.
Violence has often been described as pleasurable and entertaining. The excitement and fun which frequently characterises violence has been commented on in literature pertaining to a range of different perpetrators: drug gangs in the United States Borgios, 1989), police and urban thugs in South Africa (Malan, 1990) and soccer hooligans in England (Buford, 1991). Thus it is probably not coincidental that violence often accompanies activities associated with recreation and entertainment, such as singing, dancing and drinking. The commission of violence during or after these activities may also assist the offender to reframe his violence as entertainment, rather than destruction. These activities, and particularly alcohol intoxication, also appear to increase the potential for violence because they lower inhibitions (Australian Institute of Criminology, 1989; Campbell 1986a; Gelles 1972; Guttmacher 1967; Renvoize 1979; Vogelman 1990a). The fourth category of theories in this literature review also explores the lowering of inhibitions. These theories relate violence to crowds and groups.
1.1.4 Theories Relating Violence to Crowds and Groups
These theories examine the influences of crowds and groups on the behaviour of their members. They focus particularly on the way in which crowds and groups seem to elicit violent behaviour, even when the crowd members evidence no history of individual violence. There is thus a concern about the social psychological dynamics which operate within crowds and groups, which I will refer to as intragroup dynamics. The way in which intragroup dynamics are affected by intergroup conflict is an important dimension of intragroup dynamics, and I will address this issue at the end of this section.
1.1.4.1 Violence and the Irrationality of Crowds
Since collective violence is often misinterpreted as an act of dementia arising from crowd membership and participation, I have chosen to focus on the issue of irrationality in this section. The concern with the phenomenon of the crowd, according to Moscovici (1985), stems largely from its potential to threaten the status quo of any society. Moscovici (1985), in his book The Age of the Crowd, argues that the radical transformation of European society at the end of the nineteenth century accounts for the increased academic interest in the crowd - especially amongst conservative intellectuals who were concerned about the crowd's extremism and the revolutionary dangers that it posed.
The heightened industrialisation of Europe towards the end of the nineteenth century resulted in an increased concentration of working class men, women and children in factories which were owned by a different and wealthier class. The urbanisation that went hand in hand with this process also meant that congregations of people increased in size and number (Moscovici, 1985). The division and exploitation of this mass of people "turned the towns into battlefields where the new poor confronted the new rich" (Moscovici, 1985, p. 21). This confrontation was often manifest in strikes, sometimes instituted by new forms of organisation such as trade unions and political parties.
Moscovici asserts that these conflictual encounters were different from earlier forms of mass protest:
In those days when the mob took to the streets it was no longer to celebrate a patron saint, to take part in a carnival or stage a peasants' revolt. The people fought their masters, shouted down their not over-saintly employers and claimed what was justly theirs. (p. 21)Hobsbawm (1963) also describes the distinctive nature and expectations of these crowds:
The classical mob did not merely riot as a protest but because it expected to achieve something by its riot. It assumed that the authorities would be sensitive to its movements, and probably also that they would make some sort of immediate concession; for the "mob" was not simply a casual collection of people united for some ad hoc purpose, but in a recognised sense, a permanent entity, even though rarely permanently organized as such. (p. 111)For conservative scholars, the changing nature of crowd rebellion was hazardous to the continued "civilisation" of European society (Le Bon 1896; Sighele, 1892; Tarde 1912). Rather than analysing the social experience and social demands of the crowd, they preferred to concentrate on the crowd's internal dynamics and its ability to subordinate the individual to the collective. One of the earliest and most influential works to reflect this was The Crowd, first published in 1895 and written by Gustav Le Bon, a French doctor:
We see, then, that the disappearance of the conscious personality, the predominance of the unconscious personality, the turning by means of suggestion and contagion of feelings and ideas in an identical direction, the tendency immediately to transform the suggested ideas into acts; these we see, are the principal characteristics of the individual forming part of a crowd. He is no longer himself, but has become an automaton who has ceased to be guided by his will. (1952, p. 32)Although theories of the crowd were hotly debated, many of the French scholars agreed that the crowd was both dangerous and homogenous. In Penal Philosophy (1912) Tarde, the great French criminologist, writes:
[The crowd] is a strange phenomenon. It is a gathering of heterogeneous elements, unknown to one another, but as soon as a spark of passion, having flashed out from one of these elements, electrifies this confused mass, there takes place a sort of sudden organization, a spontaneous generation. This incoherence becomes cohesion, this noise becomes a voice and these thousand of men crowded together soon form but a single animal, a wild beast without a name, which marches to its goal with an irresistible finality. (p. 232)According to Tarde (1912) "even good men" become infected with insanity once they join a crowd. Consequently, he proposed that people who committed crimes while in a crowd should not be as severely punished as other criminals. Like many other scholars, Tarde (1892) also emphasised the animal nature of the crowd, and described it as:
An impulsive and maniacal plaything of its instincts and its mechanical habits, often an animal of the lower orders, an invertebrate, a monstrous worm whose sensibility is diffuse and who still acts with disordered movements according to the dictates of its head. (p. 358)In the face of these French writings, it was left to authors such as Emile Zola, in his novel Germinal (1921), to offer a more sympathetic view of the crowd. Although this novel, which centres around a miners' strike, does depict the savage actions of a crowd of strikers, it also portrays these actions as an understandable response to an immoral and dehumanising situation. For Zola, when men are subjected to bestial conditions, bestial retaliation is inevitable.
The views of Le Bon and other French writers were not new. Moscovici (1985) documents a long and universal list of commentaries which assert that when individuals are immersed in crowds they become unpredictable and lose their moral worth. He begins with Solon's observation that individual Athenians were as cunning as foxes, but became as "witless as a flock of sheep" (p. 14) when collectively assembled. Likewise, Frederick the Great respected his generals only as individuals, as an assembly he believed them to be "imbeciles". For the poet Grillpazer, a crowd so debased the individual that it reduced them to the status of animals (Moscovici, 1985).
Twentieth century observations are little different. Einstein (1954), in referring to the Roman proverb Senatores onmens bani viri, senatus romanus mala bestia (The senators are all good men, the Roman senate is an evil beast), wrote: " How much misery does this fact cause mankind! It is the source of wars and every kind of oppression which fills the earth with pain, sighs and bitterness" (p. 54). The Italian Marxist philosopher, Gramsci (1953), posited that the crowd, while it has the potential to be an instrument of change, is not in principle a positive agent for political transformation. Gramsci (1953) argued that a mass of people, dominated by immediate interests and vulnerable to fortuitous events, is unlikely to make correct decisions, particularly when it is composed of heterogenous elements with no social or economic ties.
Linked to the notion of the crowd as "witless" is the view that crowds reduce the intelligence of their members. Maupassant (1979), for example, comments: "How many times have I noted that when one lives alone the intelligence increases and rises, whereas once one mixes with other men once more it diminishes and falls" (p. 102). Simone Weil (1955), the French philosopher asserts: "With regard to thought, the relationship is reversed. In this connection, the individual is greater than the mass as much as something is greater than nothing, for thought only forms in a single mind face to face with itself. Collectivities do not think" (p. 108).
To digress slightly, it is worth noting that more recent social psychological theories about decision making in small groups espouse similar views. Janis's (1983) theory of "groupthink", based on an analysis of a number of foreign policy decisions made by government committees in the United States, is one such example. Janis defines groupthink as the ability of groups to ensure that individuals who would not normally do so will make an obviously wrong decision when in a group. This is more likely to occur when the group is tightly knit; well shielded from outside influences; not sufficiently exposed to alternative courses of action; has to make a decision urgently; and is under the influence of an exceedingly assertive leader. Under these conditions, the group feels omnipotent and morally superior and tends to stereotype those individuals outside of the group. Another feature of groupthink is that individuals are not permitted to express criticism or reservations about group decisions.
Research on group polarization is another area that suggests that participation in groups influences individuals and their decision making. Stoner (1961), an early pioneer of polarization research, examined decisions made by individuals before and then during group membership. His findings indicated that group opinions did not reflect the approximate average of the opinions of its constituent members, but instead were nearly always more risky than the average of the decisions that individuals made before they became members of the group. Brown (1988) points out that as Stoner's findings, and the evidence provided by other experimental studies, demonstrate that "groups seem to shift away from the 'neutral' point of a scale towards the pole which was initially favoured by the average of the individual choices" (p. 143/144), it would be more apt to describe the phenomenon as a "shift to extremity" rather than a "shift to risk" (p. 143).
While the shift to extremity has been found consistently in laboratory experiments, the evidence for this phenomenon has been less consistent in naturalistic settings where real decision-making groups have been examined (Brown, 1988). This discrepancy may be attributed to numerous influences, including the fact that in laboratory settings decisions rarely had real consequences (Brown, 1988), and that groups in laboratories, as compared to real groups, were more often transient and less likely to develop an internal structure with designated positions and established norms, all of which may curb polarization (Semin & Glendon, 1973).
There is little unanimity in the reasons proposed for group polarization. Social comparison theory, for instance (Sanders & Baron, 1977), asserts that as self image is partly the consequence of comparing oneself to others, and because individuals endeavour to maintain positive self images, one of the ways in which they can do this is to believe that their "attitudes are more extreme - in the appropriate, highly valued direction - than those of others" (Tyson, 1987, p. 348). A second explanation is the persuasive arguments theory (Burnstein & Vinokur, 1977). Polarization, according to this perspective, occurs because new arguments and information emerge during the discussions in which decisions are made. The group will finally move to the decision which is supported by the largest number of convincing arguments. Finally, the social identity approach (Wetherell, 1977) contends that polarization eventuates because group members conform to ingroup norms. For Brown (1988), all three theories can provide much supporting evidence, and all probably underpin the process of polarization.
The contention that crowds reduce rationality, reason and intelligence, sometimes leads to the assumption that individuals are less emotionally contained and tend to be more volatile when in a crowd. A theory frequently associated with this assumption, and one drawn from the ideas of Lebon (1952), is that of deindividuation. The theory put forward by Festinger, Pepitone and Newcombe (1952) referred to the loss of inner restraints which occurred when individuals were not "paid attention to as individuals" (p. 382). Subsequently deindividuation has come to generally mean a "state of reduced self awareness and concern for social evaluation" (Mann, Newtown and Innes, 1982, p. 260). Experimental social psychologists have attempted to identify a range of conditions that occur within groups and crowds that will reduce self-consciousness and a release of behaviours that were previously restrained (Dipboye, 1977). Central among these are those conditions, such as anonymity, that lessen an individual's identifiability (Zimbardo, 1969). Other deindividuating inputs that have been identified are large group size (Mullen, 1983; Zimbardo, 1969), the diffusion of responsibility (Zimbardo, 1969) and arousal (Diener, 1976; Zimbardo, 1969).
There is however, not concordance on the effect of these deindividuating inputs. Thus Diener (1976) contends, unlike Zimbardo (1969), that anonymity does not necessarily result in increases in aggression. He also argues in contradiction to Paloutzian, (1972), that findings from his experimental work show that "group presence was related to a significant decrease in aggression" (1976, p. 506). Diener's finding in this respect is consistent with other findings on the effect of groups (Diener, Westford, Dineen and Fraser, 1973). One possible reason for groups inhibiting aggression in these studies is because the groups being investigated were aggressing at low levels and aggression within such a setting would therefore have been a largely atypical behaviour. Aggression within a non-aggressive audience would heighten self-consciousness and consequently decrease the frequency of aggression (Dipboye, 1977). Thus Dipboye asserts it is more useful to understand the response to deindividuation in terms of norms that are prevalent in specific situations. Lowered self awareness, he writes has been found by Carver (1974) and Scheier, Fenigstein and Buss (1974), "either to increase or decrease aggression, depending on whether the experimenter creates a norm favouring or opposing aggression (Dipboye, 1977, p. 1071).
For those interested in maintaining the status quo, the view that crowds reduce rationality and reason provides the opportunity to dismiss any of the crowd's demands as mere folly. In this vein, popular opinion asserts that crowds that are composed predominantly of the poor and the working class, who are less educated, are even less adept at reasoning and therefore more volatile. If there is any explanation for this increased volatility however, it is unlikely that it would be based in inherent deficiencies in reasoning. Instead, it is likely to be found in the more profound experiences of discontent and frustration. The power of logical reasoning is not exclusive to anyone class - the educated may be at an advantage when it comes to reasoning, but a cursory glance at the history of human experience informs us that high levels of education are no protection against volatility, and that advanced education does not ensure a humane or non-violent approach to the world. Lifton's (1986) book, The Nazi Doctors, to which I referred earlier, provides a classic example of how some of the more brilliant minds of Europe were able to commit the most ghastly crimes in the name of science, reason and community morality.
Just as it cannot be assumed that crowds composed of the middle class or intellectuals will not be violent, it cannot be assumed that violence will be the inevitable outcome of any crowd composed of the socially discontented and frustrated. An examination of the behaviour of this type of crowd in South Africa between 1990 and 1993 highlights the varied forms of crowd behaviour. Some of these crowds were provocative and riotous, while others behaved in a disciplined and peaceful fashion that was remarkable by any international standard. The latter gatherings have sometimes involved tens of thousands of people marching through a city centre without a major police presence and without incident.
The composition, nature, and purpose of crowds vary so extensively that although it is possible to identify certain social dynamics which operate in any large group, it would be futile to put any particular stamp on the phenomenon as a whole. Canneti's (1981) Crowds and Power, although it does not sufficiently emphasise the different social interests of different crowds or the number of goals that a crowd may pursue at one time, highlights the great divergence in types of crowds. He describes baiting, flight, prohibition, reversal, and feast crowds, and with respect to rationality, he also writes of both rational and irrational crowds. As Tyson (1987) notes in his critique of Janis's groupthink theory, there are many groups who make "some very good decisions" (p. 348). Furthermore, there are crowds that are not at all brutal in their pursuit of humane and benevolent causes.
Crowds are not static, and external pressures can result in rapid and dramatic changes to their composition and conduct. Initially, the crowd may gather for "reasonable" purposes, but subsequently it may behave in a manner which contradicts these original intentions. The assumption that crowds always produce a degenerative transformation in their members also places too much culpability on the experience of being in a crowd. Crowds consist of a collection of individual histories and personalities, many of which are characterised by impetuousness and destructiveness prior to membership of the crowd.
The issue of individual histories may help to explain different individual responses to similar psychological and social pressures in a crowd. Clinical psychology, which explores the individual's subjective desires, concerns and behaviours, is particularly significant in this respect. Allport (1924) was one social psychologist who stressed the importance of examining the behaviour of the crowd in terms of its individuality. Although individuals in crowds come together for a common purpose, there may be disparate motivations for their behaviour within the crowd.
In line with much of the social psychological literature since the 1960s, I will argue that socio-psychological dynamics within groups or crowds often contribute to violence. My point of departure from Le Bon and other theorists however, is my contention that these dynamics alone are not responsible for violence. Although the crowd has its own internal dynamics, the crowd's behaviour may also be determined by primary factors such as deprivation, frustration, subcultures of violence, political organisation, and opportunity; as well as secondary factors such as physical context. With regard to the latter, academic literature and more specifically social and experimental psychology, has paid particular attention to the effects of crowding and excessive heat on aggression and collective violence.3 I will return to this subject in section 1.1.4.6, after I have examined two processes that the social psychological literature contends influence individual behaviour in crowds and groups: authorization and conformity.
1.1.4.2 Authorization
The relationship between leaders and followers in crowds and groups is a popular topic in the psychological literature (Brown, 1986; Fromm, 1974; Kelman & Lawrence, 1972; Milgram 1977; Shanab & Yahya, 1977). Leadership has been investigated in terms of whether leaders are individuals with particular personality types, or whether they are products of the situation (Bales, 1950; Brown, 1988; Sherif, Harvey, White, Hood & Sherif, 1965). Styles of leadership - dictatorial, democratic, laissez-faire - and their effects on followers have also been examined (Lippitt & White, 1943). With respect to violence, much of the interest in the interaction between leaders and followers has focused on the issue of authorization. The relationship between authorization and violence acquired particular prominence after the Second World War during the Nuremberg trials, when Nazi leaders attempted to justify their actions on the basis that they were merely carrying out orders.
Milgram (1965, 1977), in an attempt to explore the issue of authorization with regard to citizens of the United States, initiated an experiment which now occupies a prominent place in the field of psychology. Milgram's study (1965) demonstrated that authorized behaviour is more often carried out than not, and that individuals feel that there is automatic justification for authorized actions. The study assessed the amount and severity of electric shock that a subject was prepared to administer to another (a confederate of the experimenter) when ordered to do so by the experimenter. The subjects had to resolve two conflicting demands: to comply with the experimenter's instructions and to shock the victim with increasing severity, or to heed the anguished pleas of the victim (the confederate) and hence disobey the experimenter's commands. The pleas of the victim therefore acted as a counter pressure to the authorization.
Milgram (1977) expressed surprise at the findings: "The proportion of obedient subjects greatly exceeded the expectations of the experimenter and his colleagues" (p. 117). Even when the subject could hear the victim's cries of pain and pleas to stop the experiment, 62 percent of the subjects were prepared to follow the experimenter's instructions, and to apply 300 volts. Milgram's poor prediction of his subjects' behaviour was shared by other experts. In one of several studies linked to this experiment, Milgram (1977) compared the predictions of 40 psychiatrists at a leading medical school in the United States with the actual performance of subjects in the experiment, and concluded that there was a vast discrepancy between the expectations of the psychiatrists and the behaviour of the subjects.
These incorrect predictions reveal not only an underestimation of the potency of the authorization process, but an overestimation of the individual's capacity to resist this process. In order to avoid such errors of attribution, Milgram (1977) emphasizes the necessity of examining human conduct within the context of situational circumstances:
Why did the psychiatrists underestimate the level of obedience? Possibly because their predictions were based on an inadequate conception of the determinants of human action, a conception that focuses on motives in vacuo. This orientation may be entirely adequate for the repair of bruised impulses as revealed on the psychiatrist's couch, but as soon as our interest turns to action in larger settings, attention must be paid to the situations in which motives are expressed. A situation exerts an important press on the individual. It exercises constraints and may provide push. In certain circumstances, it is not so much the kind of person a man is, as the kind of situation in which he is placed, that determines his actions. (p. 118)Milgram's (1965) study did not address the significance of the finding that 38 percent of his subjects managed to resist the pressure to administer shocks, or what had enabled them to do so, and he also failed to appraise the strength of the authorization process in terms of the context and different actors involved. Different social dynamics make it easier or more difficult for individuals to give and or receive orders. For example, had there been substantial conflict and tension between the instructor and the instructee in Milgram's experiment, the instructee may have been less likely to carry out instructions. Or, if the victims in Milgram's experiment had been assigned greater status or position, the severity of the shocks administered may have decreased.
Milgram's experiment prompted a series of other studies which attempted to challenge or augment his findings (Kelman & Lawrence, 1972; Shanab & Yahya, 1977). One of the most fascinating of these was the Kelman and Lawrence (1972) study of the mass killing of Vietnamese residents in the village of My Lai by American soldiers under the command of Lieutenant Calley. This study added an interesting dimension to the issue of authorization and subcultures of violence, because according to Kelman and Lawrence the massacre did not take place as a result of a direct order, but rather because Lieutenant Calley and his men believed that the message that such actions were both required and permitted had been indirectly conveyed to them by their superior officers throughout the war.
1.1.4.3 Conformity
Conformity is another intragroup dynamic that has been linked to violence, and it has been emphasized in the literature on gangs and subcultures where violence has become normative (APA, 1993; Borgios, 1989; Buford, 1991; Siann, 1985; Vogelman, 1990a; Wolfgang & Ferracuti, 1967a). The tendency to conform has been extensively documented in the social psychological literature (Asch, 1951, 1956; Brown, 1988; Buford, 1991; Colman, Court Record; Moscovici, 1985; Sherif 1936; 1966), and two types of conformity have been identified: informational and normative. In informational conformity, people conform because they believe they lack the appropriate knowledge and that others have better information. Normative conformity, on the other hand, stems largely from the individual's desire for acceptance and belonging, and the fear of negative evaluation and social ostracism (Schneider, 1976). It is this type of conformity that is particularly relevant to this study.
Sherif (1935) and Asch (1952) undertook two of the most famous experiments that demonstrate conformity. Sherif's work centred around the autokinetic phenomenon, where a single light in the dark appears to move but is in fact motionless. Part of the phenomenon is the difficulty in assessing the extent of the light's movement. In his experiment, subjects were shown a single point of light in a dark room and then were informed that it was moving. They were then asked to approximate how much the light had moved. A confederate was then introduced who made estimates either consistently lower or consistently higher than those of the subject. The subject and the confederate then took turns in making further appraisals. The findings of this experiment indicated that with each trial, subjects made estimates increasingly similar to those of the confederate. Sherif's study suggests that individuals tend to be influenced by others when they do not believe that they have sufficient knowledge to make their own judgements. It thus demonstrates informational conformity.
In Asch's experiment, subjects were shown two cards. On the first card there were three lines of distinctly different lengths. The second card had one line marked on it, which was similar in length to one of the lines on the first card. Each subject was requested to choose the line on the first card that was the same length as the line on the second card. After a few trials, the first four subjects, who were confederates of the experimenter, all gave the same, but obviously incorrect, answer. The findings of the study revealed that 35% of the subjects chose the same answer as the confederates, even though they were aware that this answer was incorrect.
The difference between these two studies was that in Asch's study, although the answers to the questions were obvious, the subjects chose incorrect answers because of social pressure (Schellenberg, 1974). Thus it illustrated that some individuals are prepared to go along with the opinion of others despite evidence to the contrary. Moscovici (1985) said of Asch's experiment:
[It is] one of the most dramatic illustrations of conformity, of blindly going along with the group, even when the individual realizes that by doing so he turns his back on reality and truth. (p. 349)Conformity is not only a conscious act, it is frequently an unconscious pressure. Dr Andrew Colman, in his expert court evidence during the SATS trial, referred to conformity as:
An alteration in your behaviour or opinions in line with other group members, as a result of merely knowing what the other group members' behaviours or opinions are … knowing does not mean continual conscious decision- making. In the initial phase of group membership, individuals are conscious of what is expected of them, but as the life of the group continues, conforming behaviour becomes increasingly automatic and hence unconscious. (Court Record, p. 2103)In addition to the nature of social penalties, social psychological experiments have identified a multitude of variables that influence the intensity of conformity pressures, including: size of the group (Schneider, 1976); degree of cohesiveness (Freedman, O'Sears & Merrill Carlsmith, 1981; Malof & Lott, 1962; Morris & Miller, 1975) and perceived threat to the group's survival (Tajfel, 1978, 1981). The latter issue is frequently the consequence of the experience of intergroup conflict and it is to this literature, and its relevance to targets of violence and intragroup dynamics and cognitions, to which I will briefly turn.
1.1.4.4 Targets of Violence and Intergroup Conflict
Dollard et al.'s (1939) theory of displacement provides an explanation as to why the targets of violence are seldom those who are most deserving of retribution. The theory asserts that because the sources of privation are usually inaccessible and powerful, violence that results from hardship will be directed at targets that are more accessible and less powerful. In the case of men who are oppressed, this may include individuals in their homes or others in their neighbourhood who are less powerful. These victims frequently belong to the same class, racial group and geographical area as the perpetrators.
To illustrate their thesis of displacement Dollard and his colleagues made use of the concept of the outgroup, and gave as an example the persecution of the Jews in Nazi Germany. At the time of the publication of their book in 1939, this persecution had reached astounding proportions. Dollard et al. paid particular attention to two of the characteristics of an outgroup: first, an outgroup is easily differentiated from an ingroup, and second, an outgroup is perceived to be a direct threat to the ingroup. When an individual is assigned to an outgroup, the potential for him or her to become a victim of violence increases substantially.
Sherif et al's (1965) group norm theory did much to advance the concepts of in-and outgroups and their implications for intergroup conflict. This theory posits "that dislike of the out-group is merely the other side of the coin of attraction to the in-group" (Buss, 1973, p. 586). Members of the ingroup assume that the traits of the outgroup are opposite to their own, and because the ingroup believes that they embody positive traits, the outgroup is assumed to possess negative traits. Sherif et al.'s (1965) study of American boys at a summer camp, demonstrated that even when individuals are assigned to groups for obviously superficial reasons, individuals develop distinctive and more positive beliefs about their group as compared to members of other groups. These attributions are even more exaggerated if the conflict is such that only one group can be victorious.
According to Tajfel (1981), the stereotyping of traits serves a number of important functions. At a cognitive level, it helps us to organise our understanding of the world (Allport, 1954; Brown, 1988). Stereotyping introduces a "simplicity and order where there is complexity and random variation" (Tajfel, 1981, p. 132). At a social level, stereotyping contributes to the "creation and maintenance of group" ideologies, explaining and justifying a variety of social actions" (Tajfel, 1981, p. 146). Furthermore, it helps to "preserve or create positively-valued differentiation of a group from other social groups" (Tajfel, 1981, p. 146). Thus stereotyping of groups and the categorisation of these groups as outgroups helps those within the ingroup to maintain a positive sense of themselves, and to maintain a high level of antagonism towards those designated as the outgroup.
1.1.4.5 Dehumanisation
Categorising an individual as a member of an outgroup is one of many ways of devaluing their worth. This devaluation is frequently referred to in the psychological and sociological literature as dehumanisation (Kelman, 1973; Litewka, 1977). Dehumanisation is dependant on the definition of the victim as the "other" so that he or she is perceived of as different and as belonging to another, less worthy, category of people. The definition of "other" can assume diverse forms. For example, it may be framed in sexual terms, as is frequently the case in rape, where the victim becomes a "slut" and a "whore" (Vogelman, 1990a). In the South African political scenario violent repression by the apartheid government against the black community was facilitated by racism - one of the more popular forms of dehumanisation both in and outside of South Africa.
Wolfgang and Ferracuti (1967a) and Guttmacher (1967) contend that certain forms of socialisation encourage the dehumanisation of others and therefore make it easier for individuals to perpetrate violence. They base this assertion on the premise that the quality and style of childrearing influences the child's regard for human life. As a result, those children who have not been treated with respect have little respect for both their own lives and the lives of others. Guttmacher (1967) writes:
One can safely hypothesise that the amount of satisfying nurture that the child receives in its earliest years must be a fundamental element in the formation of attitudes on the value of human life. (p. 121)While this view has some validity, it is challenged by the fact that many individuals who have received good enough nurturing and have had enormous social opportunity and privilege still behave in a brutal manner. Such individuals frequently behave with enormous compassion and care toward those people that matter to them - their families, their religious grouping or their community - while all of those outside of their circle are painted with the brush of "other" to a lesser or greater degree, and so are seen to deserve the violence that is directed at them. It is thus vital to look to the realms of political traditions, social conflict and the internalisation of ideologies in order to fully understand the process of dehumanisation.
I will conclude the section on crowds and groups with a brief review of two physical factors which may help to promote violence: overcrowding and temperature.
1.1.4.6 Physical Influences on Crowd and Group Violence
1.1.4.6.1 Overcrowding
The view that overcrowding has a detrimental effect on social behaviour is partly based on the work of biologists and ethologists such as Lorenz (1966) and Leyhausen (1965). Studies using animal subjects revealed that high population density led to a breakdown in normal social conduct and an increase in aggression. These findings were, however, contested on the basis that they cannot be generalised to humans. Dubos (1970) writes:
The readiness with which man adapts to potentially dangerous situations makes it unwise to apply directly to human life the results of experiments designed to test the acute effects of crowding on animals. (p. 207)Experiments conducted with human subjects have not really provided any clarification because of the inconsistency of their results (Freedman, 1972). The tentative conclusion that can be drawn from such experiments is that overcrowding tends to reinforce existing feelings and emotions, and that the extent of this effect is dependent on a number of variables, such as: duration of exposure; amount of activity required; the formality and informality of the situation; the expectations of the individual (Freedman, 1972); and the sex of the subjects (Siann, 1985). With respect to the latter, for instance, Siann (1985) notes that some research indicates that crowding results in higher levels of aggression in males than it does in females, although the reasons for this are unclear.
1.1.4.6.2 Temperature
The relationship between weather, climate and violence has been examined at length by criminologists. The Italian criminologist Sighele (1892), and the French criminologist Tarde (1892) believed that riots were influenced by both weather and climate. They maintained that uprisings rarely took place during cold or rainy weather.
The relationship between temperature and violence received much attention as a consequence of the civil violence in the United States in the 1960s and 1970s. The United States Riot Commission (1968) and many social psychologists implicated hot and humid weather conditions in the outbreak of violence. The Commission reported that during all but one of the violent disturbances in 1967, the temperature was above 80 degrees fahrenheit (27 degrees celsius). Baron and Ransberger (1978) investigated 102 cases of collective violence between 1967 and 1971, and found that "the frequency of riots was related to temperature in a curvilinear fashion, with an inflection point in the 81-85 fahrenheit (27-29 celsius) interval, well above the average annual temperature in the USA" (p. 159).
These studies associated heightened temperatures with violence, because increased temperature is believed to increase irritability and arousal. Berkowitz (1972), for instance, in his work on frustration, asserts that temperature exacerbates the experience of aversion and thus contributes to violence. He describes the role of temperature in the violence in the United States in the 1960s and 1970s as follows:
For blacks exposed to all too many social and economic frustrations in their daily lives anyway, this weather would be yet another unpleasant experience, one more occurrence inflaming nerves and creating a tinderbox to be ignited by some spark. (p. 80)Tyson and Turnbull (1990), in what is probably the only in depth study on collective violence and temperature in South Africa, attempted to assess the applicability of North American findings to South Africa. Their findings indicate that the mean celsius temperature of "riot days" in South Africa was 22,4, compared with 26,10 in the United States. Furthermore, the mean temperature for incidents of collective violence in South Africa was not significantly different from the country's annual mean temperature of 22 degrees celsius. They concluded that "riots [in South Africa] did not occur on days that were unusually hot" (p. 160). Four different categories of theories have been outlined thus far: biological, developmental, cultural, and those linking violence to crowds and groups. The fifth category that I will consider consists of those theories that relate violence to inadequate political, economic and social conditions and structures.
1.1.5 Privation Theories
This approach contends that oppressive conditions, structures and institutions within a society are a major determinant of violence - be it political, social or criminal (APA, 1983; Gurr, 1968a, 1970; Libman-Rubinstein, 1979; Lodge, 1983; Wilson & Ramphele, 1989). The conditions and structural dimensions most commonly referred to include a lack of a political democracy, political persecution, poverty, unemployment, socio-economic inequality and police violence. Klineberg (1981) writes for instance, that "there is more violence when there is no generally positive attitude towards the legitimacy of the government, [and] a feeling that it is not responsive to popular [democratic] needs and wishes"(p. 6/7).
This approach also stresses that a belief in violence is, in part, borne of acute frustration and a loss of hope in the efficacy of non-violent protest (Gurr, 1979; Libman-Rubinstein, 1979). Libman-Rubinstein (1979), studied collective violence perpetrated by different groups, such as Native Americans, Ku Klux Klan, labour, and civil, and different historical periods - 1790s, 1850s, 1890s and 1930s - in the United States. According to this study:
Most groups which engaged in mass violence have done so only after a long period of fruitless, relatively non-violent struggle in which established procedures were tried and found unavailing. (p. 143)Gurr (1979) makes a similar point regarding civil violence in the United States in the 1960s:
The desire for change inspires protest in the hope that groups in power will respond favourably, while despair about resistance is often translated into rebellion. (p. 50)As I have indicated, this approach associates violence with the discontent arising from political, economic and social factors (APA, 1993; Cock, 1990; Gurr, 1979; Klineberg, 1981). However, it is also necessary to acknowledge that an individual may experience dissatisfaction in one life sphere, but continue to experience sufficient satisfaction in other spheres to assuage this discontent. Thus an individual who endures discrimination, but has the opportunity to acquire wealth and live peacefully, may experience more satisfaction than dissatisfaction. In this study, however, I will focus not on this situation, but on the experience of widespread grievances and a generalised discontent with life conditions.
The danger of inferring a causal relationship between onerous conditions and discontent is highlighted by the example of the oppression of women. Some feminists, for instance, incorrectly assume that all women who live in sexist cultures or who encounter sexist practices are discontented (Ettorre, 1982; Lewis, 1982). However, there are many women who regularly confront what Western liberal standards would term sexist practices and yet do not appear to feel victimised or dissatisfied. Religious women are a case in point. Ironically, these women may actually experience more contentment than those women who are critical of sexist practices. This satisfaction may stem from the internalisation of patriarchal orthodox religious ideology, which asserts that sexist practices are natural and godly, or alternatively from the positive sense of self which is acquired through the traditionally defined role. Whatever the reason, many of these women do not view themselves as oppressed.
This point is particularly significant with respect to economic hardship, which is often causally related to violence (Songer, 1916; Hovland & Sears, 1940; Mintz, 1946; Shaw & MacKay, 1931). One major study that discounts such a causal relationship is Snyder and Tilly's (1972) Hardship and collective violence in France, 1830 to 1960. Their results show no significant correlation between collective violence and economic hardship, instead, they found that levels of collective violence were determined largely by the degree of conflict over resources, amount of political activity and the extent of state repression. Snyder and Tilly (1972), in their explanation of collective violence, place great emphasis on the political struggle between conflicting interest groups. They assert that collective violence "tends to occur when one group lays claim to a set of resources, and at least one other group resists that claim" (p. 526). Thus Snyder and Tilly (1972) refer to prospects resisted and impeded, a concept closely related to Dollard et al.'s (1939) understanding of frustration. The work of Snyder and Tilly (1972, 1974) also served as a major challenge to relative deprivation theorists such as Gurr (1968), who interpreted collective violence primarily as a response to a gap between expectations and achievement.
1.1.6 Relative Deprivation Theories
Relative deprivation theory has made a seminal contribution to the literature in that it highlights the importance of the expected and substantive changes in political, economic and social conditions, and the influence of these on the generation of discontent and violence. Ted Gurr is the most prominent of the politically orientated relative deprivation theorists. In his book Why Men Rebel (1970), he contends that social dissatisfaction, and the rebellion that frequently follows, is partly the consequence of the gap between expectations and achievement. According to Gurr and other researchers such as Davies (1974), Cook, Crosby and Hennigan (1977), and Folger, Rosenfield and Rheaume (1983), the discrepancy between what an individual achieves, and what he considers his due, is a major cause of discontent. Consequently, the experience of austere conditions is worsened when there are expectations of an improved future that are not realised. Gurr (1968a) asserts that the importance of relative deprivation is such that it is not merely a magnifying factor, but rather the fundamental precondition for civil strife.
For North American social scientists, unmet expectations were a favoured exposition for the civil violence in the United States in the 1960s and 1970s. This concept was also used to explain the perplexing finding that blacks in the northern states, who had experienced better conditions and less overt racism than their southern counterparts, rioted more often (Berkowitz 1972; Caplan, 1970; Sears & McConahay 1970a). Sears and McConahay (1970a) argue that social violence frequently occurs when people are making progress in improving their conditions, because in these contexts expectations are likely to rise faster than achievements (Schneider, 1976).
Davies' (1974) J-curve theory of revolution has its own special place in relative deprivation theory. According to Davies, revolution and political violence transpire when prolonged economic and social development is followed by short sharp reversals. This dramatic negative change in living circumstance, Davies asserts, is responsible for fuelling desperation and resentment. However, although such experiences of rapid change are likely to heighten dissatisfaction, it is fallacious to assume that this dissatisfaction will automatically lead to violence. One of the most obvious and notable exceptions to Davies' thesis is the great depression of 1929 to 1932 in the United States. This depression followed a period of rapid economic growth in the 1920s, and yet during the depression little political violence and no revolution took place (Eckstein, 1980). Thus while falling standards of living may play a crucial role in establishing the potential for revolution and violence, numerous contingent factors, such as strong political organisation and a weak, illegitimate government, are often necessary before any tumultuous events can take place. Snyder and Tilly (1972), who maintain that the roots of collective violence lie largely in the struggles for political power and resources, make this point in their critical evaluation of Gurr, Davies and other "expectation theorists".
Psychologists Hyman (1942), Merton and Kitt (1950), and Festinger (1954) have contributed to the theory of relative deprivation with the formulation of the thesis of social comparison. According to Festinger (1942) and others (Lewin, Dembo, Festinger & Sears, 1944), levels of expectation are determined not only by past successes or failures, but also by the human drive to evaluate oneself against others, and more specifically against similar others.4 Thus relative deprivation theory contends that an individual's level of contentment is influenced in three ways: by the extent to which their expectations have been realised (Gurr, 1970); by means of a comparison of their present and past circumstances (Davies, 1974); and through a comparison of their own circumstances with those of similar others (Festinger, 1954).
2. Conceptual Framework: The Nine Phase Social Psychological Model of Violence
The social science literature has traditionally analysed violence through the use of either a psychological or a sociological orientation. The psychological viewpoint emphasizes those factors associated with the individual's development and functioning, and has often explained violence in terms of psychopathology, inferior personality attributes, irrationality or vulnerability (Le Bon, 1896; Lombroso, 1911a; Tarde, 1892), while ignoring the impact of political, economic and social forces. In the sociological analyses, the reverse is true (Davies, 1974; Snyder & Tilly, 1972).
The richest accounts of violence, I believe, require an integrated understanding of all the major components of human life: the biological, the psychological and the sociological. As I have not examined the contribution that neurophysiology may have made to the Eight's conduct, a major limitation of this study is the exclusion of the biological. At this point it is important to note the work of Fromm (1974), as he has managed to incorporate all three elements with some success. Fromm's work, however, is concerned more with individual destructiveness and human nature and the building of psycho-analytic theory than it is with comprehending collective violence. Within the social psychological domain there are numerous studies which straddle both the psychological and sociological, but these studies often fail to provide an integrated perspective of the aetiology of violence because they focus on the analysis of a specific contributing factor or the validation of a particular hypothesis. Straker (1992), for example, is primarily interested in the answer to the question "does violence beget violence?" (p. 87), while Gurr (1970) is focused on proving his hypothesis that relative deprivation helps to impel political violence. Such work is advantageous because it explores a specific factor, which potentially contributes to violence, in depth, but it fails to incorporate the host of other factors that are crucial to an understanding of the aetiology of violence.
Buford (1991) and Malan (1990) provide two more notable examples of work that incorporates both the psychological and sociological. The former study deals with soccer hooliganism in England, and the latter with specific incidents of violence in South Africa. Both are brilliant expositions of violence, but they are journalistic in character and are more concerned with description than with developing a framework through which violence can be understood. Thus I maintain that the literature does not evidence a commitment to developing an integrated approach to causality which considers the multitude of causes for violence and the ways in which they intersect. This study, and the model I have developed, is an attempt to reflect such a commitment.
This study will focus on the subjective histories of the Eight as well as their past and present social circumstances, and their violence will be understood as a product of the interface between the subjective and the objective. I will view violence not as the outcome of one to one causal relationships, but rather as the product of the interconnection between a host of substantive and mediating factors. This perspective is similar to those studies of collective violence that view the phenomenon as an attempt by collectives to re-shape their social worlds, and by individuals to re-fashion an improved sense of self (Fanon, 1968; Gurr, 1968, 1970; Perry & Pugh, 1978). The authorities' understanding of such attempts will also be seen as a key determinant in deciding whether they attempt to ameliorate the perceived injustices, whether they too use violence, and whether this in turn leads to an increase or decrease in the violence committed by the subjugated.
Within dialectical conceptualisations of behaviour new ways of being arid relating are continually being established or expressed, and while particular values and behaviours may normalise over time they are never permanent. Changing situational circumstances, fluctuating power relations, shifting perceptions of law enforcement agencies, new opportunities, and desensitisation may all have a dramatic effect on the forms that collective violence assumes. Even when violence becomes normalised, the form it takes often changes continually. This social dialectic, combined with human creativity, probably accounts for the consistent ability of human beings to devise new ways to maim and kill each other.
Although I have argued that no single theory is able to explain collective violence adequately, new theories are necessary to enhance our understanding of the behaviour. This study aims to contribute to the dialogue between theories by expounding a new conceptual model, based on the extension of some of the basic concepts of an old one: the frustration-aggression hypothesis. The model should not be viewed as an attempt to devise a single theory of violence, as the disparate nature of violence makes this an impossibility. The development of social science is premised on theories that stimulate each other, and I am hopeful that in the same way as my insights have been fuelled by some of the theoretical propositions of the frustration-aggression hypothesis, my model too will contribute towards an understanding of other incidents of collective violence and the behaviour of other violent offenders.
The frustration-aggression hypothesis originated in the seminal work of Dollard, Doob, Miller, Mower and Sears (1939), collectively referred to as the Yale Group. This group, and the revisionist explorations of their work in the United States in the 1960s and 1970s (Berkowitz, 1972; Caplan, 1970; Caplan & Paige, 1968; Gurr, 1968; Perry & Pugh, 1978), attempted to decipher the violence of their times. Dollard et al.'s (1939) book, Frustration and Aggression, explained both criminal and fascist violent behaviours in terms of a mechanical connection between frustration and aggression:
This study takes as its point of departure the assumption that aggression is always the consequence of frustration. More specifically the proposition is that the occurrence of aggression always presupposes the existence of frustration and contrariwise, that the existence of frustration always leads to some form of aggression. (p. 1) Frustration was referred to by Dollard et al. as "a condition which exists when a goal response suffers interference" (p. 11).Frustration is therefore an interference preventing the realisation of an expected gratification. Dollard et al. detail numerous types of interferences. They write: "such expressions as 'to disappoint a person', 'to let someone down', 'to cause pain to someone' and to 'block someone in carrying out an act' indicate that one person is imposing a frustration on another" (Dollard et al., 1939, p. 7). Unlike Dollard et al., who view frustration as embracing a myriad of different behaviours, this study will employ a more narrow understanding of frustration, and one that is particularly pertinent to conflict between authorities and collectives. Frustration will thus refer to discontent that stems from the unsuccessful attempt to rectify the conditions that are perceived to be at the root of individual or group dissatisfaction.
Dollard et al.'s (1939) definition of aggression is contradictory. Initially, they define aggression in behavioural terms as any "sequence of behaviour, the goal response to which is the injury of the person to whom it was directed" (p. 9). Later, however, they refer to aggression in ideational terms, it: "may exist as the content of a phantasy or dream or even a well thought-out plan of revenge" (p. 10). Dollard et al. also fail to distinguish between anger and aggression - a distinction that is relevant to this study. In this study, anger refers to an emotion characterised by extreme displeasure, whereas aggression is any behavioural manifestation of anger where there is an intent to do harm.
Before proceeding to a critique of other aspects of Dollard et al.'s work, it is necessary to highlight some of the significant conclusions of their study. According to Dollard et al., the extent to which frustration affects the strength of an aggressive response is dependent on: the strength of the drive for the expected gratification; the degree of interference with the attempt to obtain the expected gratification; and "the number of frustrated response sequences" (Dollard et al., 1939, p. 28). As I will do in my model, Dollard et al. indicate that every frustration does not necessarily lead to overt aggression towards the source of the frustration, because the perpetrator of aggression may want to avoid retaliation. Aggression, however, does not necessarily disappear, but can be displaced onto more vulnerable targets.
Another notable conclusion of the study, but one upon which they do not concentrate, is the finding that frustration may occur after aggression. Dollard et al. write of a "vicious circle - frustration, aggression, interference with aggression, more frustration - [that] tends to be repeated as long as successive acts of aggression suffer interference" (p. 40). This point is often overlooked in commentaries on the frustration-aggression hypothesis (Australian Institute of Criminology, 1989; Hoffmann & McKendrick, 1990; Klineberg, 1981), possibly because Dollard et al. do not elaborate on this conclusion. In my model however, the notion, of aggression as a potential antecedent for frustration is a key one.
Although the mechanistic approach adopted by Dollard et al. was supported by a multitude of experiments designed to demonstrate the causal relationship between frustration and aggression (Siann, 1985), it was also criticised extensively (Maslow, 1941).5 In response to these criticisms, Miller (1941) revised the thesis by introducing the proposition that although there may be non-aggressive responses to frustration, aggression is the dominant response to frustration, and frustration increases the potential for aggression.
Another limitation of the frustration-aggression hypothesis is that it fails to consider the possibility that aggression may occur not because people "have been thwarted in the past, but because they think this action will bring them some other benefits" (Berkowitz, 1989, p. 62). For instance, the aggression of the police may at times have little to do with frustration, and more to do with individual quests for power, promotion and social status. The latter aggression is referred to in the social psychological literature as instrumental aggression because unlike hostile aggression, its primary aim is not to inflict harm (Berkowitz, 1989; Feshbach, 1964). This distinction was not utilised in this study, as the aggression and the violence committed by the Eight appeared to be both hostile and instrumental.
Scholarly work is inclined to respond to and reflect social developments, rather than to initiate them. Much of the violence in the United States in the 1960s and 1970s was a response to inequitable civil rights, discrimination and poverty, and as a result, researchers during this period added a central component to the frustration-aggression hypothesis: the concept of deprivation, which was defined mainly in terms of socio-economic hardship. Many of the articles written about violence during this period continued to identify additional contingent variables that were believed to influence the nature of the relationship between frustration and violence (Abeles, 1976; Berkowitz, 1972; Blau & Blau, 1982; Caplan, 1970; Caplan & Paige, 1968a; Cohen, 1970; Flango & Sherbenou, 1976; Gurr, 1968, 1970, 1979). In view of the prominence of the deprivation-frustration-aggression (DFA) thesis, and its relevance to any understanding of collective violence, Gurr (1968), one of the great scholars of violence, remarked that the thesis had acquired the status of a fundamental social law. There have been numerous other subsequent challenges and amendments to the frustration-aggression thesis, but the premise that frustration, however it is defined, plays an integral part in aggression and violent acts remains widely accepted.
Although this study is more concerned with violence than with strikes, it is important to note that prominent authors within the realm of strike theory do not view the frustration-aggression hypothesis as particularly helpful in building an aetiology of strikes. Bluen (1994) for instance, excludes the relevance of the hypothesis because he states that Dollard et al. referred primarily to hostile aggression, and that strikes typically involve instrumental rather than hostile aggression. Kelly and Nicholson (1980) contend that given the disparate aims of strikes, Dollard et al.'s hypothesis might, at best, predict only a "small number of one-off wildcat strikes, arising from unexpectedly blocked work goals and satisfactions" (p. 865). While these criticisms of the frustration-aggression hypothesis may be justified, they may be also less applicable if revisions in definition and concepts to the frustration-aggression hypothesis are accepted.
Despite these and other criticisms of the frustration-aggression hypothesis (Bandura, 1973; Baron, 1977; Zillman, 1979), I have found some of the concepts delineated in the hypothesis to be valuable. Most significantly, the theory highlights the strong connection between aversive experiences and frustration and aggression. It can refer to both individual and collective experiences, desires and goals and can be generalised to encounters at an interpersonal, familial, political and occupational level. The hypothesis was also particularly pertinent to this study because, as I will illustrate, the Eight suffered from a number of blocked opportunities and much of their behaviour was characterised by an intent to do harm. Thus although the frustration-aggression hypothesis in its archetypal form is not relevant to this study, to every strike or to all forms of violence, a reformulated version of the hypothesis is useful in explicating the conduct of the Eight.
In addition to definitional changes, my reformulation of Dollard et al.'s theory and the DFA framework is based on the formulation of nine phases which trace the evolution of violence. Within these phases I have included three dimensions which I believe are essential to any analysis of violence: factors which pertain to the individual, group norms and dynamics and the social context. The advantage of the inclusion of different phases in the model is that it serves to emphasise the passage of time that is necessary for the emergence of violent behaviour. Within these phases, I will indicate a host of other non-violent behaviours which occur during the evolution of violence, and I will also explore the reasons why violence is perceived as preferable to non-violence in dealing with certain targets.
Before I outline this model, I need to preface it with a number of academic vindications. It is exceedingly unlikely that the model is applicable to all forms of collective violence, as it was developed specifically to explain the violence in which the Eight participated. The model attempts to delineate the reasons for the occurrence of violence, but it does not explain why, in certain situations where violence would be expected to occur, it does not. The model is also primarily focused on explaining the occurrence of the ultimate form of violence: murder.
My model was not developed as a hypothesis to be validated by evidence gathered from the Eight and other sources. Instead, it originated in an attempt to provide a coherent explanation of the data accumulated on the Eight's violence. While my model attempts to identify the antecedent conditions which contribute to collective violence, it does not endeavour to understand the primary source of this violence. With regard to the Eight's violence, apartheid will be identified as a major contributor, but I will not attempt to explore the origins of this system in seventeenth century European colonialism. I contend, but do not explain, that the roots of the strikers' violence in 1987 are deep and extend well back into the preceding centuries. Similarly, there is insufficient scope to analyse in detail some of the more immediate antecedent conditions of their collective violence - such as striking. At this point it should be noted that because a labour dispute was pivotal in the collective violence that is examined in this study, many of the examples that I will use to illustrate my argument will refer to the occupational domain.
Although the nine phases of my model are described in a linear fashion, and are clearly differentiated for purposes of clarity and emphasis, in practical terms they intersect and overlap. No one phase exists in isolation, and each phase incorporates the residue of the dynamics of earlier phases. The different phases are distinguishable principally by the predominance of new attitudes, emotions, behaviours and social pressures. Thus although certain forms of violence may have emerged at the outset of the strike, they are only described in phase five because it was at this point that they emerged with intensity.
Because the phases interact in a cyclical pattern, there is no fixed time period before a new phase materializes. It may be days, weeks, months or years. New phases emerge as a result of an interaction of a number of factors, including: the intensification of emotions, a change in the dynamics between actors, the search for more successful strategies and tactics, and the influence that the disappearance or the emergence of new conditions brings to bear on the actors concerned. The model is therefore an open system with every phase conditioned "by external tensions and strategic considerations and factors that focus awareness or defuse conflict" (Kelly & Nicholson, 1980, p. 868).
I have adopted Eckstein's (1980) concepts of inherent and contingent conditions to explain the varied import and interaction of factors which contribute to violence. Inherent conditions do not cause collective violence, rather they are the rudimentary conditions that underpin this violence. In order for violence to materialise, these inherent conditions must intersect with a host of contingent variables. These variables, which are accidental, uncertain, and peculiar to time and place, provide the impetus for the progression from non-violence to violence. Contingencies thus build on inherent conditions to hinder or facilitate violence, and in themselves they affect the nature, intensity and form of each phase. In reality, contingent and inherent conditions are intertwined and difficult to disentangle, but for the purposes of clarity I have attempted to separate the two and I have included only the inherent conditions in phase one.
Before I outline the model, it must be noted that the Eight's experiences – which provide the substantive evidence for the phases - will not be chronicled in this chapter, but from chapter four onwards.
2.1 Phase One: Hardship
As privation theorists (Arendt, 1970; Caplan, 1970; Caplan & Paige, 1968a; Fanon, 1968; Geary, 1986; Gurr, 1968, 1970, 1979; Rude, 1964) might have predicted, the collective violence of the Eight was founded in austere and cruel conditions in the political, economic and social domains. The conditions I refer to encompass poverty, wholly inadequate work conditions, repression, severe human rights abuses, and the absence of democratic rights. This study is not so much concerned with an exact assessment of the nature of each of these conditions, for example a conclusive definition of poverty,6 but rather with the Eight's experience of and dissatisfaction with obviously harsh conditions. In essence, I will show that these conditions caused dissatisfaction because they were responsible for inflicting both physical and psychological harm. The form of this harm varied according to the hardship endured, for example, those who were racially abused were subject to humiliation and those with uncomfortable living conditions were frequently subject to sleep difficulties.
Harsh conditions, whether they result from discrimination, low wages or poor living conditions, do not automatically generate significant discontent as Perry and Pugh (1978) and others (Gurr 1970) point out. In this study, however, I will focus not on the situation where the discriminated and oppressed appear to be satisfied, but on the experience of widespread grievances and a generalised discontent with life conditions.
Although this study focuses on harsh conditions in the political, economic and social spheres, I will also refer to the effects of abusive conditions in the family on the individual. As is the case with political, economic and social hardship, and excluding those encounters which cause obvious physical pain, the child's interpretation of what others may classify as demeaning experiences will be influenced by his or her own expectations and knowledge of alternatives (Widom, 1989). Because harsh conditions do not necessarily precipitate the experience of discontent, they obviously cannot presuppose violence. Furthermore even where discontent exists, violence cannot be assumed to a logical consequence of this experience. Snyder and Tilly (1972, 1974) and Libman-Rubinstein (1979) highlight the myriad of mediating and contingent variables that need to be considered in any explanation of how and why violence emerges in response to hardship. These variables are crucial in determining whether dissent and frustration become violent, and they will be explored in more detail in section 2.7.
2.2 Phase Two: Deprivation
Phase two of my model reflects the Eight's sense of loss, and their view of their life as unsatisfactory because of their experience of harsh conditions. My model thus characterises deprivation as both objective and real hardship, and incorporates the component of discontent about this hardship into this definition. The quest to thwart the damage caused by harsh political, economic and social conditions has, as I indicated in my literature review, prompted a broad range of disciplines - including clinical psychology, political science and sociology - to view deprivation as the basis of political and civil violence (Caplan, 1970; Fanon, 1968; Gurr, 1968; Rude, 1964; Straker, 1992).
Discontent is not a mechanical response to inequitable conditions, but it may arise as a consequence of a number of contingent factors. In cases where feelings of discontent were present at the outset, contingent factors will augment this dissatisfaction. The sections which follow will examine two key contingent factors: politicisation and relative deprivation.
2.2.1 Politicisation
Politicisation is an influential factor in encouraging individuals to interpret their experience of harsh conditions in a negative light (Fanon, 1968; Gurr, 1970; Piven & Cloward 1977; Swilling, 1987). In the broadest sense, and with reference to the deprived, politicisation is any process that informs individuals that they have the right to better conditions; that their plight is the result of a particular social system and its power relations; and that there are preferable social alternatives. Piven and Cloward (1977) contend that active resistance and protest movements emerge when new found political consciousness includes both the belief that the individual is entitled to better conditions, and also the perception that social systems are not immutable but have the potential to be changed. The latter insight can, as de Tocqueville (1856) has observed, heighten discontent considerably. In explaining the roots of the mass scale resentment which led to the French Revolution, de Tocqueville (1856) wrote that evil "patiently endured … comes to appear intolerable once the possibility of removing it crosses men's minds" (p. 176-177).
Politicisation, as I will demonstrate, can take place in a number of different ways, including casual conversations (Perry & Pugh, 1978), exposure to political rhetoric (Reich, 1948), songs (Sherman, 1989) and exposure to the media (Posel, 1989). However, it would seem that the process, is at its most powerful when the individual becomes an active participant in a movement that supports his political, economic and social interests (Coser, 1974; Perry & Pugh, 1978; Swilling, 1987). Two other factors which can enhance the politicisation process are the authorities' increasing loss of credibility (Piven & Cloward, 1977) and the knowledge that others who were similarly deprived in the past have been successful in their struggle (Fanon, 1968).
2.2.2 Relative Deprivation
The theories of relative deprivation provide the second set of contingent factors that significantly influence the experience of discontent. I have discussed these theories in the literature review, and I will illustrate their relevance in the case of the Eight in chapters four and five. The common thread that unites the different relative deprivation theorists is the contention that discontent resulting from deprivation is promoted or amplified by negative experiences of change and adverse observations of difference. Relative deprivation theorists focus on negative changes in the individual's material existence (Davies, 1974), improvements that are anticipated but never realised (Cook, Crosby & Hennigan 1977; Folger, Rosenfield & Rheaume 1983; Gurr, 1970; Sears & McConahay 1970a), and negative social comparisons that are made by the individual (Festinger, 1954; Hyman, 1942; Merton & Kitt 1950).
2.3 Phase Three: Anger and Circumspect Non-Violent Protest
Sadness and anger are two principal feelings associated with loss and dissatisfaction (Bettelheim & Rosenfeld, 1993). When sadness overwhelms the individual, the result is often despondency, despair and passivity (American Psychiatric Association, 1987; Rohrbaugh, 1981), and because sadness is one of the most painful emotional states, individuals seek numerous methods to avoid or minimise this state - in my clinical experience these have ranged from alcohol intoxication to overwork. One emotion which is extremely effective in keeping sadness and other negative feelings at bay is anger that is directed at an external target (Gilligan, 1991). Anger directed at the self leads to self deprecation and poor self esteem, both of which are sources of sadness (Bettelheim, 1986). Externally directed anger, on the other hand, assists the individual to avoid internal reflection, the breeding ground of sadness. According to Fitch (1970), this kind of anger also bolsters self esteem, because others, as opposed to the self, are blamed for unsatisfactory conditions.
Anger is not merely a psychological defense against sadness, it also stems from the quest to improve one's life conditions and to protect oneself from further abuse. When anger is exhibited in behaviour which is characterised by objection and disapproval directed at the source of the dissatisfaction, I have termed this protest. More often than not, protest is non-violent, and in those situations where it is violent, it is likely to have been preceded by non-violent protests (Gurr, 1979; Libman-Rubinstein, 1979). The reasons for the shift from non-violent to violent protest are an important focus of this model (see section 2.7).
Anger is also frequently related to conflict, because it may be related to the desire to punish and hurt those responsible for inflicting suffering, and to restore the imbalance that has led to feelings of degradation and powerlessness (Arendt, 1970; Fromm, 1974). Fromm (1974) has termed the latter function the quest for existential equality. Where the manifestation of anger includes this intent to do harm, I have termed it aggression.
In this phase of the model, protest is non-violent in character and expressed in an economical and circumspect manner. In the case of the Eight, this was reflected in the manner in which the Eight protested against racist insults - their resistance was never combative and was couched in a way that would not offend the racist (see chapter four). Their prudence, as Dollard et al. (1939) would probably suggest, appears to have been related to the ability of those responsible for their privations to exact penalties for protest and aggression. The greater the perception of this power, the more cautious the deprived are likely to be in their behaviour. Conversely, if those in authority are seen to lack the capacity for retaliation and censure, then protest and aggression may progress to violence more quickly (see section 2.7.1.4.1).
Protest may remain informal, or it may progress to more formal institutions that are designed to facilitate peaceful conflict resolution. This progression is partly dependent on the availability, credibility (Garson & O'Brien, 1979; Gurr, 1979; Snyder & Tilly, 1972) and accessibility of institutions and channels for non-violent grievance resolution (Johnson, 1966). In many cases, this is not an all or nothing scenario. For example, in apartheid South Africa although blacks were disenfranchised and did not have the option of resolving their grievances through formal government institutions such as parliament, occasionally they were able to defend their interests or exact retribution through the courts. However, because the courts comprised largely of racist judges and magistrates, and cases were adjudicated on the basis of apartheid legislation, there were few instances of success. Consequently, they had little credibility amongst the black population, and were regarded as a last resort.
In addition to availability and credibility, longevity and historical standing also play a part in determining whether the deprived will make use of institutions. Garson and O'Brien (1979) make this point with respect to the high incidence of violence after the civil war in the United States. They assert that institutional mechanisms of non-violent conflict resolution that were instituted after the conclusion of the war were not trusted or used because they were newly established. Consequently, they had little restraining influence.
Where the avenues for non-violent conflict resolution are problematic, individuals may articulate their discontent informally in discussions with colleagues, friends or family. In the course of these conversations, the individual ascertains whether or not his anger is shared and justified (Piven & Cloward, 1977; Seekings, 1991). When the individual believes that it is both, the potential for collective aggression and non-violent protest action against the perceived source of physical and psychological damage increases dramatically (Berkowitz, 1989; Perry & Pugh, 1978; Piven & Cloward, 1977).
2.4 Phase Four: Frustration
Earlier I indicated that my definition of frustration is different to that adopted by Dollard et al. in that I view frustration as a discontent that is a consequence of an unsuccessful protest to change the conditions that give rise to discontent. Expectation theory proposes that the stronger the desire for change, the higher the resulting levels of discontent if change does not occur (Dollard et al., 1939, Gurr, 1970; Cook, Crosby & Hennigan, 1977). In contrast, when the deprived are hopeful that non-violent protest or aggression will be successful, discontent is diminished (Dollard et al., 1939). Governments and other authorities appear to be aware that this optimism has the potential to restrain violent behaviour. Accordingly, when authorities are placed under severe pressure, a common response is to promise to investigate and re-evaluate conditions, or to pledge to rectify them. The less sophisticated approach, and one of the quickest methods of extinguishing hope, is to respond to non-violent dissent with violent repression. This, as McKendrick and Hoffmann (1990) point out, is the response that most frequently leads to the desire for violent retributive justice and counter-violence.
2.4.1 Frustration and the Intensification of Deprivation
The relationship between deprivation and frustration is not unidirectional, as the ineffectuality of protest not only gives rise to frustration, it also prolongs and often intensifies deprivation. Frustration fuels deprivation because it emphasizes feelings of loss and powerlessness. The interaction between frustration and deprivation is clearly evident in cases where the authorities subvert protest through physically and financially harmful penalties such as imprisonment and dismissal, as was habitually the case in apartheid South Africa. These practices may then lead to further protests in relation not only to poor conditions, but also to the nature of the penalties invoked by the authorities. These protests too may be blocked, further increasing the levels of frustration and deprivation. This cycle is likely to repeat itself many times, and with every repetition the experiences of deprivation and frustration are intensified.
Deprivation and frustration are often transformed into demands that are expressed through more militant forms of protest through the catalyst of a trigger event (Kelly & Nicholson, 1980). Triggers may be seemingly minor events, but they may also, as this study will demonstrate (see chapter five), be major occurrences in themselves. The progression towards violence and murder may be strewn with a number of trigger events - in this model, there is a trigger to militant aggression, a trigger to violence and a number of triggers to murder.
2.5 Phase Five: Militant Aggression
The difference between politely informing a supervisor that one is displeased with a racist remark, and openly going on strike, highlights the distinction between circumspect non-violent protest and militant aggression. The former is judicious and reflects a tempering and restraining of anger. The latter is more combative, overt and active. Both actions exhibit a challenge to authority, but militant aggression presents its challenge in a more open and straightforward manner. In this study, the strike reflected an act of militant aggression because its aim was to hurt SATS management as part of an attempt to improve conditions. The presumption was that the employer, in an attempt to avoid the costs of militant aggression, would rectify the grievances which gave rise to such aggression. In this sense, militant aggression incorporates both hostile and instrumental aggression.
Numerous factors determine whether militant aggression will follow frustration. Buss (1961) asserts that aggression is unlikely to follow frustration "unless it is accompanied by the knowledge that aggression will be useful in ridding oneself of frustration" (p. 140). Conversely, if the consequences of militant aggression are perceived to be too negative (Dollard et al., 1939), then frustration may lead to withdrawal and hopelessness as opposed to aggression. Klandermans (1984, 1986) describes the latter process in his value expectancy theory of strike causation (see chapter five). He argues that an evaluation of the personal costs and benefits of the action is central to any worker's decision to strike. Another factor that is considered in chapters five and seven, and that influences the choice between retreat or militant aggression in response to frustration, is the extent to which individuals are prepared to make sacrifices for a cause (Arendt, 1970; Fanon, 1968; Gilligan, 1991)
When militant aggression does manifest, its form is mediated by factors such as the extent to which forms of non-violent circumspect protest have been exhausted; the political and economic climate; the militancy of attitudes; collective identification (Kelly & Kelly, 1992); and where and how individuals gather (Smelser, 1962). If militant aggression is exhibited in the workplace, the important mediating factors include the size and nature of the work organisation (Kerr & Siegal, 1954); managerial style (Kelly & Nicholson, 1980) and communication patterns in the organisation (Kerr & Siegal, 1954; Walton & McKersie, 1965; Whyte, 1951); the existence of a trade union; whether the trade union, if there is one, is moderate or militant; and the nature and skills of the strike leadership or instigators of protest (Batstone, Boraston & Frenkel, 1978; Hyman, 1977; Turner 1968).
2.6 Phase Six: Frustration Despite the Use of Militant Aggression and the Continuation and Intensification of Deprivation
If militant aggression is not successful in securing a change in the conditions of deprivation, frustration is likely to follow. This frustration may be particularly intense, as Gurr (1970) points out, if there were strong expectations that the militant aggressive protest would be successful. The response of the source of deprivation to the militant aggressive protest is also significant. Often, because a protest action is militant, the response of those in authority is equally combative and may involve mass dismissals, violence and killings. Consequently, previous frustrations may now be accompanied by additional forms of deprivation.
If frustration is viewed as a possible consequence of militant aggression, and not merely as an antecedent, the question of the validity of the catharsis theory of aggression arises. The concept of catharsis is rooted in the writings of Aristotle, who contended that the audience of Greek tragic theatre, in vicariously experiencing emotions like fear and pity, purged themselves of these emotions (Siann, 1985). Freud (1930, 1932, 1933) related this concept to aggression and argued that in order to be psychologically healthy, the aggressive instinct required some form of release. Aggressive behaviour was one way of providing this catharsis for the aggressive instinct. Subsequently, Dollard et al. (1939), Lorenz (1969) and Storr (1970) all emphasised the advantageous effects of expressing aggression.
Later in this study I will examine in more detail, and by way of example of the Eight, the reasons why I propose that militant aggression and violence do not always have sufficient cathartic effect to reduce further aggression and violence if deprivation persists. In contrast, the catharsis of aggression and violence may be so appealing that it generates the desire to repeat the behaviours over and over again.
Earlier in this chapter I explained that within the framework of this model the emergence of violence follows a trigger event. In chapter five I will indicate that this trigger was a substantial event which did much to heighten the Eight's already intense deprivation and frustration.
2.7 Phase Seven: Violence
Although it is impossible to predict the exact point at which violence will take place, it is possible to predict whether violence will manifest within a broader time period. The early warning signs for collective violence are evident in the combination of heightened deprivation and frustration, and the host of contingent conditions which create a strong desire and opportunity for individuals and collectives to commit violence. In the process of outlining some of these contingent conditions, it will be argued that they also influence the form of the violence that ensues.
2.7.1 Contingent Conditions that Facilitate the Choice of Violent Behaviour
2.7.1.1 The Lack of Alternatives and Desperation
In order for an individual to choose violence as a means of conflict resolution and the achievement of goals, there must be a belief in its efficacy (see section 2.7.1.2) (Arendt, 1970). The perpetrator is also likely to feel despondent about the capacity of non-violence to successfuly ameliorate deprivation (Gurr, 1979; Libman-Rubinstein, 1979). Because frustration presupposes the continuation of deprivation it frequently also embodies, as was illustrated with the Eight, a bleak perception of the future (see chapters seven and eight). This outlook can foster the belief that any action which will alter the likelihood of a dismal future is worthwhile, and this perception in turn encourages daring and bold behaviour.
The availability of alternatives and resources which cushion the impact of continued or worsening deprivation plays an important role in determining the extent of individual or group desperation (Billig, 1976; Siann, 1985). In the political realm, these alternatives may include the opportunity to seek refuge in another place or country; while within the realm of labour disputes, they include the strikers' access to union strike funds or dismissed workers' access to social welfare and unemployment benefits.
Biologically inclined theorists such as Lorenz (1969), assert that the extremity of the competition for resources is dependent on how essential these resources are to physiological survival. In conflict where defeat has radical implications, and given the failure of non-violent change mechanisms, the deprived are often prepared to go to extremes to ensure victory (Segal, 1991b; Webster & Simpson, 1990).
2.7.1.2 Belief in Violence
A belief in violence may originate in a number of different experiences of violence: as a victim and witness to actual displays of violence (Hoffmann & McKendrick, 1990; Pizzey, 1974; Renvoize, 1979; Vogelman 1990a), or as a viewer of enacted violence, namely media violence (Friedrich-Cofer & Huston, 1986; Lefkowitz, Eron, Walder & Huesmann, 1972). In general, these experiences teach an individual that violence provides power to the offender and leads to the victim's submission. The locale for these lessons of violence may be the home, the family, the school, and the political and social arenas (APA, 1993; Hoffmann & McKendrick, 1990).
2.7.1.3 Risk and Self Sacrifice
Although acute frustration may engender pessimism about the possibility of an improved future, individuals do not necessarily lose all hope and they continue to be concerned about the quality of their lives (Bettelheim, 1986; Levi, 1988). Consequently, although many individuals who are desperate and frustrated may be prepared to incur greater personal risk for the sake of change, it seems that few would be prepared to engage in violence if they were aware that the outcome could be social ostracism, long term imprisonment or death (Dollard et al., 1939; Vogelman & Eagle, 1991). Nevertheless, there are some individuals whose commitment to a particular social or political cause overshadows any personal risk. Fanon (1968) provides an example of the latter scenario, with the assertion that the colonised are prepared to sacrifice self preservation for dignity. While Fanon's assertions have some validity, particularly in the case of hunger strikers and suicide bombers, it would seem that for most individuals the reality is, as Arendt (1970) points out, that self preservation is at least as important, if not more so, as the conservation of dignity.7
Even where there are high risks involved and individuals do engage in violent resistance, they may do so within the realms of measured recklessness. This implies that there is always the belief that there is some chance, no matter how small, of escaping negative consequences. As I will demonstrate in the case of the Eight, perceptions of law enforcement and the nature and power of group morality are important influences in an individual's decision to commit violence.
2.7.1.4 Legal Consequences, Detection and Opportunity
A multitude of factors influence an individual's perceptions of the possibility of escaping detection and judicial consequences, as well as the form that the violence will take. Some of these factors are described below.
2.7.1.4.1 Effectiveness of law and order agencies and the judiciary
The effectiveness of law and order agencies and the judiciary, in terms of the success rates of arrest, prosecution and conviction of offenders, is influenced by police training methods, the extent of reporting and public assistance (itself partly influenced by credibility and perceptions of effectiveness of the police and the judicial system), personnel resources and access to appropriate technology (APA, 1993; Vogelman, 1990b; Vogelman & Eagle, 1991). The potential offender's perceptions of the strength and potency of law enforcement agencies and the judiciary influence his degree of confidence in committing acts of violence (Geary, 1986; Snyder & Tilly, 1972; Vogelman, 1990a; Vogelman & Eagle, 1991).
2.7.1.4.2 Geography of the scene of the offence
Perceptions of the safety of the venue may be influenced by the degree to which the venue is concealed from public view and the way in which the structure of the venue makes it easier or more difficult for those wanting to prevent violence to succeed in doing so.
2.7.1.4.3 Transport facilities
The types of transport that are available to potential offenders, for example access to vehicles, may enable them to choose venues which are more conducive to concealment and escape.
2.7.1.4.4 Time of day
Certain times of the day offer better opportunities to commit certain types of violence. For instance, in South Africa massacres which involve a number of perpetrators and victims - and so have to retain