Centre for the Study of Violence and Reconciliation

Nongoloza's Children:
Western Cape prison gangs
during and after apartheid

Jonny Steinberg

Monograph written for the Centre for the Study of Violence and Reconciliation, July 2004.

Jonny Steinberg is an independent consultant.



Chapter 1: Nongoloza and Kilikijan

Chapter 2: The functions of violence I—Making men (and not children)

Chapter 3: The functions of violence II—Making men (and not women)

Chapter 4: Prison on the streets, the streets in prison

Chapter 5: Warders and gangs



Over the past two decades, news of the strange world behind the bars of South Africa's prisons has been spilling out in dribs and drabs. Among the things we have learned is that the so-called "Number gangs"—the 26s, 27s and 28s—are about 100 years old, that they originated in the jails, mine compounds and informal settlements of turn-of-the-century Johannesburg, and that today they constitute a formidable force in every prison across the country. We know that the Number gangs take their inspiration from the real historical figure who founded them, Nongoloza Mathebula, an early Johannesburg bandit who built a quasi-military band of outlaws, welding his small army together with a simple but potent ideology of banditry-as-anti-colonial-resistance. We know, too, that the Number gangs have been the vehicle of an extraordinarily durable oral tradition; the imaginary uniforms, weapons and paraphernalia that Number gangsters carry today are all faithful representations of the uniforms, weapons and paraphernalia of the Boer and British armies of the late 19th-century Transvaal. The arcane and finely observed military and judicial hierarchies of the 28s and the 27s are precisely those invented by Nongoloza and described in the life testimony he dictated to a prison warder in 1912.1

We know, too, that the world of the Number gangs is one of staggering brutality. Its self-styled judiciaries sentence inmates to death, to gang rape, to beatings with prison mugs, padlocks and bars of soap; among the prerequisites of joining the "soldier lines" of the gangs is the taking of a warder's or a non-gangster's blood; leaving a prison gang, sharing a gang's secrets with a warder, or talking casually about the gang's workings to the non-initiated are all punishable crimes.

Finally, we know that sexual relations between prison gangsters and their lovers are highly stylised, caricaturing the most pungently misogynist relationship imaginable between a man and a woman. The passive partner in an archetypal prison relationship is stripped of the jail equivalent of his juridical personhood: he is not allowed to conduct commerce or to leave the cell without his partner's permission; he cooks for his partner, makes his bed, washes his back and cuts his toenails.2

Yet, despite the growing body of information available to us, our knowledge of South African prison gangs remains inadequate. What we have are slivers of narrative, ritual and myth, disconnected from the context that gives them meaning. What we do not have is an analysis of the relationship between prison gangs and the institution that animates them in the first place: the prison itself.

This lacuna has been unavoidable. Until the early 1990s, South African prisons were entirely closed institutions. Reportage on the conditions in jails was effectively illegal until the mid-1980s. Until 1990, the courts kept prison administration at arm's length, giving the Commissioner of Prisons almost unlimited power, including the power to regulate information.3 And the gangs themselves were, of course, sworn to a vow of silence.

Two research projects conducted in the 1980s did, however, make concerted and valuable attempts to understand prisons gangs in their institutional context. Fink Haysom's thesis—that the gangs' quasi-military structure is an extreme parody of the apartheid prison system itself—is, as we shall see, a powerful and lasting insight.4 Lötter and Schurink's thesis—that gangs militate against the psychological and material pains of imprisonment—is also of lasting value.5 However, given the closed conditions of the 1980s, these studies, which should have become a foundation for further research, were left standing as two islands in a sea of incomprehension.

It is only in recent years, with the opening of the jails to researchers and observers, and the intrusion of the rule of law in the running of prisons, that it has become possible to study the substance of prison life with any seriousness. And it is only in understanding the substance of prison life that the strange narratives and rituals of South Africa's prison gangs can possibly make sense.

Aims of the monograph

This monograph is by no means a comprehensive history of prison gangs in the Western Cape. Rather, it is potted and thematic, its conclusions interpretive and discursive. It is hoped that other researchers might use its ideas as a basis for further work. The aim of what follows is to provide a modest addition to our understanding of prison gangs in the following areas:

A word on method

This monograph rests on two legs of research. First, I conducted nine months of fieldwork at Pollsmoor Prison's Admission Centre between the beginning of October 2002 and the end of June 2003. During that time, the prison held more than 3 000 inmates, slightly more than two-thirds of whom were awaiting trial. I spent half of each day shadowing a senior warder as he did his work. I spent the other half interviewing inmates, usually informally, sometimes with a set of written questions and a tape recorder. I was not allowed into the prison during lockup—4pm to 7am—the period of each day when the presence of warders contracts to a skeletal staff and the Number gangs take effective control of life behind the locked cell doors.

Second, over an 18-month period I conducted intensive interviews with about 30 veteran Number members. All had served time in prisons across the Western Cape during the 1980s and 1990s. The oldest among them provided detailed testimony of prison life in the late 1970s.

As is immediately apparent, my access to prison life in the past and prison life in the present came in two very different forms. In regard to the past, I was relying mainly on other people's memories. (The gangs obviously keep no written record of their histories, and the records of prison authorities, judicial commissions and court trials are far too partial, fragmentary and inaccurate to stand in as documented history.) Memory is a poor witness to the past in the best of circumstances, and these were hardly the best of circumstances. As you will see later, there is a deep and angry divide between veteran Western Cape prison gangsters and those who cut their teeth in the 1990s and 2000s. Veterans speak endlessly of how the transition from apartheid has corrupted the vital traditions of the Number, of how the gangs today are disfigured wreckages of their former selves. Veterans invariably tell idealised stories about the past. They do so, not simply to deride what the current generation has done to the Number, but also because imprisonment is a humiliating, often harrowing, experience, and constructing one's life narrative before a witness/researcher inevitably involves restoring dignity. The interviewee mutes the acts of violence he has committed and often omits the violence committed against him. The world of the 1970s and 1980s that emerges from the mouths of veterans is invariably sanitised.

I thus found myself embroiled in an endless adventure of corroboration: tracking down franse (non-gangsters) who had witnessed events related to me by gangsters; contrasting a 28's version of a particular event with that of a 26. Sometimes I was successful, sometimes not; inevitably, some of the idealisations of my interviewees have found their way into this monograph.

My access to prison life in the present was also not ideal. The politics of any prison is shaped by a daily cycle. Between 7am and 4pm, inmates are free to walk around their sections, to wash clothes, to mingle, occasionally to exercise in the yard. They do so under the gaze of warders, a gaze that surely shapes their behaviour. This was the time segment to which I had access.

For the rest of the time—15 hours of every day—between 40 and 60 men are locked in a cell designed to house 18. As I argue later, it is the deprivations suffered in this situation that account or much of the logic of prison gang narrative and prison violence. My sense of what happens during this 15-hour segment of every day relied on the testimony of inmates, on my own imagination, and on inductions based on what happened during open time.

Finally, while my research into the past elicited testimony from literally every prison in the Western Cape, my research into the present was restricted to Pollsmoor Admission Centre. Pollsmoor is the railway station of Western Cape prisons. It is situated within a stone's throw of Cape Town's ghettos; most awaiting-trial prisoners in the greater Cape Town area are processed there. There is a great deal of flux, of movement, of exchanges between the streets and the prison. The jails of the Western Cape hinterland are quieter, more stable.

Chapter 1: Nongoloza and Kilikijan

The historical Nongoloza

The 26s, 27s and 28s all originate from bands of outlaws that plagued late 19th and early 20th-century Johannesburg. The largest and most memorable of these gangs was called The Ninevites; its rank and file were lumpenproletarians—young black men who had left their ancestral land in the countryside but had refused to take up wage employment for white bosses in the early mining town.

The Ninevites were led by a charismatic young Zulu migrant, "Nongoloza" Mathebula. (He was born Mzuzephi Mathebula. Nongoloza is the name he adopted at the height of his underworld reign.) Imbued with a crisp and feisty imagination, which had been instilled by the injustices that lay in his own past, Nongoloza shaped his crew of outlaws into a paramilitary hierarchy. It borrowed its rank structure and its imaginary uniforms from the Natal colony's judiciary and the Transvaal republic's military. Perhaps most interesting of all, Nongoloza imbued his bandit army with a political purpose. "I reorganised my gang of robbers," he reported to his white captors in 1912. "I laid them under what has since become known as Nineveh law. I read in the Bible about the great state Nineveh which rebelled against the Lord and I selected that name for my gang as rebels against the Government's laws."6

The Ninevites lasted nearly two decades. At their height, in the early 1900s, they boasted a membership of nearly 1 000. They launched their sorties into robbery and plunder from a series of caves and warrens that stretched across the south-western perimeter of Johannesburg; they also gained effective control of the inmate populations of several of the mining compounds and prisons of early Johannesburg. Had they chosen to practise their profession differently, they might have been remembered as an African pedigree of Robin Hoods, taking from the victors of South Africa's colonial order and giving to the downtrodden. As it was, the Ninevites showed little discernment when it came to their choice of victims. Among their favourite pastimes was to rob black labourers as they made their way home on payday.

And so early Johannesburg's black proletarians remembered Nongoloza with a mixture of fear and awe. It was said that he and his bandits established an underground world in a disused mineshaft, replete with shops, beautiful white women and a Scottish bookkeeper.7

It was also said that Nongoloza himself was imbued with magic, that the bullets of white policemen and soldiers bounced off his skin.8 In early proletarian lore, he was something of a Janus-faced monster: horrible because he was undiscerningly brutal, enticing because he showed that "even the poor can be terrible".9

The Ninevites were crushed in the mid-1910s. Nongoloza himself, extraordinarily enough, renounced his gang and agreed to work for the prison authorities. But by then, most of the gang's leaders had spent time in jail and had begun to recruit there. Thousands of young black men, criminalised by white South Africa's racial laws, drifted in and out of the prisons of early 20th-century South Africa.10 By the early 1930s, gang derivatives of the Ninevites had a presence in almost every prison across the country. They have been there ever since, the memory of Nongoloza and the legends of his life passed down from one generation of prisoners to the next, throughout the 20th century.

It is quite extraordinary how much of Nongoloza's imagination has been preserved in the prison gangs of today—the 26s, 27s and 28s. The imaginary uniforms copied from the early Boer republic are still there. So are the imaginary .303 rifles and bayonets that the Boer commandos took into battle with the British in 1899. Nongoloza's original rank structure, dividing members between soldiers and judicial officers, and dividing the judicial officers themselves between an upper and a lower court, is still extant.

Most interesting of all, the Number gangs have held onto the mainstay of Nongoloza's original ideology. All three are organised around a largely mythical narrative of the great bandit's career. Indeed, they place the origin of their own division into three rival gangs in Nongoloza's times. And yet, while they disagree about episodes in his life, and about decisions he made in regard to the nature of banditry, all agree that he became a bandit because blacks were being disinherited of their land and forced to work like slaves in the mines. In other words, throughout the century, South Africa's prisons have incubated a fiercely anti-colonial ideology.

The mythical Nongoloza

What follows is an account of contemporary prison gangs' mythical narrative of Nongoloza's life. It is cobbled together from interviews with about three dozen Western Cape prisoners and former prisoners. Any Western Cape prison gangster who reads it will quibble with at least part of it. As with all oral histories, there are countless variations in the story. More than that, there are rival versions, each allied to a competing doctrinal position; the 26s, 27s and 28s disagree about certain things Nongoloza thought and did, and about decisions he made. Aligned to this disagreement about the story, there is disagreement about practice in the here and now, about how prisoners ought to live their lives in 2004.

The story begins in an African village somewhere in South Africa, on the brink of industrialisation. The 28s say it is a Zulu village. The 27s don't specify where it is, but they insist the village is not Zulu, anything but Zulu. There is an elderly man: the 28s call him Ngulugut, the 27s call him Pomabasa, or Po. He is a wise man and a seer, and he embodies the interests of all black people.

During Po's autumnal years, the young men of his village begin leaving their kraals to look for work in the gold mines of the new city that the whites have built. Oddly, time and place have changed. The year is 1812, 74 years before gold was discovered on the Witwatersrand. And the place of the gold mines is Delagoa Bay, on the northern-most reach of the eastern coastline, not Johannesburg.

The young men leave their kraals for the gold mines, but they never return. And so the seer, wondering what has happened to the bearers of his village's future, journeys to the mines himself. He spends time in the single-sex mine compounds where the young black men stay, and he soon discovers why they do not return. The work beneath the ground is not fit for brutes; the young men are dying as they dig up the white men's gold.

So Po flees Delagoa Bay and retreats to a cave somewhere on the outskirts of the town of Pietermaritzburg and gives thought about what to do. His cave is agter die berge (behind the mountains) a place of solitude and contemplation. He spends the first weeks in his retreat inventing a secret language, for he knows that if the young men are to be saved, the whites must not understand the talk between the men who are to become his followers.

Po's cave is a short distance from a perch. Sitting at the perch one day, he looks out over the roads that lead from the hinterland to the mining town. One morning he sees a cloud of dust on the road that comes from Zululand. He descends from his lair, goes out onto the road, and finds a young man in the cloud of dust. He asks the stranger his name. "Nongoloza," the young Zulu replies.

Po asks him where he is going.

"To the mines," Nongoloza answers, "to look for work."

The old man shakes his head. "I have been to the mines," he advises, "and I have seen what happens there. The work will kill you in the years to come."

Nongoloza asks the wise man what he should do instead.

"The gold of the white man is good," Po replies. "You must take it, but not from the ground. You must rob it from the white man himself."

Po takes Nongoloza up into his cave and the following morning, sitting at his perch, he sees another cloud of dust, this one on the road from Pondoland. The events of the previous day are repeated, but this time the name of the youngster on the road is Kilikijan and he is a Pondo. Po entices him up to his cave, and so things go on until Po has gathered 15 young migrants around him. He instructs them in the secret language he has invented, he tells them of the pay wagons that roll into the mine compounds on Fridays and he teaches them the art of highway robbery.

The young bandits are successful at stealing wages but, holed out as they are in their cave, they need other provisions as well, like food and clothes. So Po directs them to attack the colonial army camps that mark the perimeter of the mining town, and this they do. As with the pay wagons, they have success. In addition to pillaging food and supplies, they also bring back with them the accoutrements of warfare: .303 rifles, bayonets, army uniforms and the rank structure of the colonial military.

By now, Po's men are wanted and hunted. The whites advertise rewards for their capture. They must change their ways to avoid detection. So they become nomads, moving from cave to warren, using the hills outside the mining town as their camouflage. They also divide themselves into two groups. Kilikijan takes seven men and robs by day. Nongoloza takes six men and robs by night. For a long time, working in this way, they terrorise the whites, taking their gold and hounding their army.

Since the beginning, Po has been instructing Nongoloza and Kilikijan to keep a diary. There is a large rock in the vicinity of one of the caves to which the men periodically retreat. He has told them to inscribe their activities as bandits onto the rock, to record how they go about their business and live their lives. This they have done. The rock is covered with the record of their short history as outlaws.

Po now brings Nongoloza and Kilikijan together and instructs them to go to a white farmer called Rabie. He tells them that they must buy a bull that grazes on Rabie's farm, and he is very specific in his instruction: there is a particular bull the men must bring back—its name is Rooiland.

The two bandits arrive at Rabie's front door and offer to pay for Rooiland. The white man is suspicious. He has heard of the bandits who are roving the outskirts of town. He refuses to sell his bull and instructs the men to leave his property. But they will not go until they have carried out Po's order. So they kill the farmer with the bayonets they have plundered from the white army. They find Rooiland in Rabie's fields and herd him back to the cave where Po is waiting.

The 15 bandits throw a tremendous feast as they slaughter Rooiland. Po presides over the slaughter. He tells the men to preserve particular parts of the beast: the hooves, the legs, the eyes, the ears, the tail and, Po says, more important than anything else: the bull's hide.

Once the animal has been dismembered, Po calls Nongoloza and Kilikijan to his side. He tells them to take one of Rooiland's horns and to fill it with a mixture of the beast's blood and gall. He then instructs both men to drink from the horn.

Kilikijan is the first to drink from the horn. He grimaces, spits it out and then exclaims: "There is poison, in here; this stuff will kill me." Then Nongoloza takes a sip, swallows it and smiles. Kilikijan stares at him in horror: the Zulu drinks poison.

Once the drinking rituals have been completed and Kilikijan has discovered that his bandit brother drinks poison, Po instructs the two men to take the hide, drape it over the rock on which the band's diary is inscribed, and press it against the rock, until the diaries are imprinted on the animal's skin. The words of the diary, now duplicated—one on the rock, the other on the hide—are to become the law of the gang. Whenever there is a dispute about what bandits ought to do, Po says, consult the hide or the rock, because they are a record of how things were done at the beginning, and how things ought to be done in the future. Nongoloza rolls up the hide and takes it with him. Kilikijan is left with the rock.

As you might expect, this business of the duplication of the record of the law was soon to cause trouble. The old rock on which the diaries were written was large and cumbersome. Kilikijan's outlaw band was always on the move, and had to take the wretched thing wherever they went. One day, high up in the hills, one of the rock's carriers stumbled and it rolled down the side of a valley. Somewhere on the slopes, the boulder crashed into a tree and broke in half. The part of the rock that hit the tree imprinted its content on the bark. The rest rolled down into the Moliva River and floated downstream to be lost forever. Kilikijan made his way down to the tree, peeled off the bark and took it with him. The bandit was now in possession of only half the law. The rest had drowned in the stream. Nongoloza, though, possessed the whole law, for he had Rooiland's hide.

At some point after the rock had been lost, the two bands—Nongoloza's and Kilikijan's—went out pillaging together. I am not sure why. The usual practice was for Kilikijan to work by day, Nongoloza by night. Just as the bandits were about to leave their hideout, Nongoloza announced that he was ill and wanted to rest. He asked that one of Kilikijan's men, a youngster called Magubane, stay behind to tend to him. So 13 went out to plunder and two stayed behind.

Kilikijan returned during the course of the afternoon to stumble upon Nongoloza making love to Magubane under a cowhide. Incensed, he raised his sabre and told Nongoloza to get up and fight. Nongoloza demurred. He said it was written in the law that what he and Magubane were doing was permitted. It says on Rooiland's hide, Nongoloza explained, that women are poison and that soldiers must choose wives from the young men in their ranks.

This only enraged Kilikijan more, since the bandit had only half the law in his possession. He could never know what was written on the original hide and what Nongoloza added later. Indeed, in years to come, the 27s were to deny that there ever was a hide; according to them it was invented retrospectively by self-interested sodomites.

So Kilikijan took a swipe at Nongoloza with his sabre and the two men fought until, it is said, Nongoloza was ankle-deep and Kilikijan knee-deep in blood. Po, who had come down from his lair at the sound of the clashing sabres, appeared on the scene. Horrified that the two bandits had hurt one another, he ordered them to put down their weapons and enquired about their dispute.

Being a sage and a seer, the old man did not resolve the disagreement in a simple manner. Instead, he issued the terms of a riddle. He told Kilikijan to go the mine compounds in Delagoa Bay to see if sex between men was practised there. He refused to be drawn, though, on the significance of the meaning of Kilikijan's findings, whatever they might be. If the gold miners did indeed have sex with one another, what precisely would this mean? He did not say.

Po also said something else. Should the two bands of men ever return to his lair, they would find a rock at its entrance. Under the rock they would find an assegai. If the blade of the assegai was rusted, it would mean that Po was dead. Upon entering his cave, they would find his skeleton. Needless to say, that is precisely what came to pass. Neither of the bandits was to see the old man alive again. The adjudicator of the law went to his death without ever pronouncing on the legitimacy of sex between bandits.

Having listened to Po's instructions, Nongoloza and Kilikijan went their separate ways. Kilikijan went to Delagoa Bay to enquire about the sexual practices of the miners. He left Magubane behind; his band was now composed of seven men. Nongoloza headed to a place called Germiston. He took Magubane with him; his band now had eight men. That is one of the explanations for the numbers 28 and 27. There are others, but this one was repeated to me most often. I am not sure where the "2" in 28 and 27 comes from. Perhaps there were 55 bandits rather than 15. Or perhaps the "2" signifies the two original bandits: Kilikijan and Nongoloza.

In any event, the two men were never to talk of their dispute again. They next met several years later in the cells of Point Prison, Durban. Both had been captured and tried for their crimes; they had been given indeterminate sentences and faced the prospect of spending the rest of their lives in jail.

If the two pioneers never again spoke to one another of their dispute, their respective followers talked about it all the time—quietly, and among themselves—for generations, spreading rumours and casting aspersions on the other camp. Both camps agree that Kilikijan did, in the interim, find his way to the mine compounds in Delagoa Bay, and that he discovered that the men there did indeed sleep with one another. The 28s say that this vindicates Nongoloza's statement that the legitimacy of sex between bandits had always been written on the hide. The 27s disagree. Kilikijan's discovery, they argue, only confirmed that sex between men is a foreign practice, one alien to those initiated into bandit life. It arose at the compounds because the work underground was so hard: weak men needed help from stronger men with their picks and shovels, and they gave them sex in exchange. Sex between men is a pollution, a symptom of the unnaturalness of the work white men forced blacks to perform.

The dispute is, of course, incapable of resolution. The rock on which bandit law was written has long been eroded by the waters of the Moliva River. And all that remains of Po, the sage, is a mute skeleton. In years to come, this irresolvable quarrel is to shape relations between the gangs. It is bound up with the question of why Nongoloza was able to drink poisonous gall and smile. And it was to frame the immediate events that arose in those first weeks when Kilikijan and Nongoloza found themselves together in Point Prison.

There were six inmates in Point Prison. They were voëls or birdies, that is to say, members of neither camp, non-gangsters, franse, nothings, who sat in a circle and flipped a silver coin. Their leader was a man by the name of Grey. Among the first practices to emerge among the 27s and 28s of the prisons was the confiscation of the possessions of the franse. A portion would always be returned, but it was to be the ndotas, the Number men, who determined the distribution. When Nongoloza's men demanded from Grey that the six inmates hand over their possessions, coin and all, he refused. Troubled by this disobedience, Nongoloza approached Kilikijan and asked what was up with these recalcitrant men. Kilikijan, who had arrived in prison before Nongoloza, explained that the flipping of the coin was a form of gambling, that these men were trained in the arts of smuggling and acquiring valuables.

During his first days in prison, Kilikijan continued, he had stabbed a troublesome warder. As punishment, he was placed in a tiny dungeon and was fed a spare, saltless diet. The warders' aim was to make him weak. The six gamblers, led by Grey, who was skilled in the art of smuggling and had the cunning to enter into prudent allegiances, had slipped him salt and bread and other nourishment under the crack beneath his door.

Impressed and curious, Nongoloza asked Kilikijan to bring him Grey's coin. He handled it carefully, then bit it, then dropped it on the floor. "It is hard, like a nail—a spyker," Nongoloza said, "and when it drops to the ground it makes the noise of a nail. I will call it a spyker. It will be useful in the years to come. I can use it to button my uniform."

"It is called a kroon (crown)," Kilikijan replied, "not a spyker. It is useful because it brings wealth."

The disagreement between the two men extended beyond the question of what names to give things. 27s and 28s offer rival versions of what happened next. According to the 27s, the two men fought over the six gamblers. While the stakes were never openly spoken of, it was quite clear what was going on: Kilikijan wanted to protect Grey and his colleagues from the appetites of the sodomites. Absorbed into the 28s, they would be used, he believed, not just to smuggle, but for sex as well.

According to the 28s, there was no such dispute. They say that Nongoloza said to Kilikijan: "I give you permission to constitute these men as the third camp of bandits, but on several conditions. First, they will be called the 26s, to show that they will never rise above us. Second, they will be the last camp to form. There will never be a fourth camp in this prison. Every other inmate is a frans. And, finally, you will be responsible for their conduct. You will be answerable for them. When they commit a wrong, I will not go to them, I will come to you."

"That is all well and good," Kilikijan replied. "But when you wrong them, I will come to you."

And so the three camps were formed, each with their self-made philosophies of banditry and their collectively assigned roles. The 26s were to accumulate wealth, which was to be distributed among all three camps, and acquired through cunning and trickery, never through violence. The 28s, in turn, were to fight on behalf of all three camps for better conditions for inmates. They would also be permitted to have sex, in their own ritualised manner, among themselves. They were never to touch a 26.

As for the 27s, they were the guarantor of gang law; they were to keep the peace between the three camps. They would learn and retain the laws of all three gangs, as well as the laws of the relationships between gangs. And they would right wrongs by exacting revenge: when blood was spilled, they would spill blood in turn.

Today, in 2004, that is how South Africa's three major prison gangs understand their origins. In the 26s and the 27s, sex between gang members is formally outlawed and subject to severe and violent punishment. Although, as you will soon see, this ban on sex is breached all the time, and the ways in which it is breached are interesting.

The 28s, in contrast, are divided into two parallel hierarchies, two lines. There is the gold, or gazi, line which is the military line and consists of soldiers who fight the gang's battles. At the apex of the gold line is its first ancestor, Nongoloza. Then there is the silver, or private, line, which is female. At the apex of this line stands its first ancestor, Magubane.

The meanings of the Nongoloza myth

The manner in which I have told the Nongoloza myth—as a single narrative, relayed by a speaker (or in this case, a writer) to an audience—is somewhat misleading. In prison, the Nongoloza myth is seldom told. Rather, it is woven into the very fabric of gang practice. Every gang-related interaction, every judicial sitting, every meeting, is described and enacted through metaphor, and the metaphor itself is drawn from the Nongoloza myth. In other words, the Nongoloza myth is not a story one tells, but a set of practices one enacts.11

For instance, in the 28s, Rooiland's carcass is the sacred object around which the structure of the gang is shaped. The emblem of the 28s is a Zulu shield, and the shield's skin is Rooiland's hide. Each substructure of the gang consists of four men and is named, respectively, "the four points of the ones", "the four points of the twos" and "the four points of the threes". The four points refer to the places in Rooiland's carcass where the legs join the body. The function of every officer in the gang is signified by a part of the bull's anatomy. The magistrate is given the hooves: they are his four stamps—red, green, white and black—with which he marks a member for promotion or punishment. The gwenza gets the legs: they are his four pens—also red, green, white and black—with which he notes the record of every member. The glas, who communicates the gang's decisions, gets one of the horns: it is to be his bugle, with which he announces the conclusions to the gang's deliberations. The draad, the gang's intelligence officer, is given the eyes, signifying that he sees all that happens.

In short, every time a meeting of the 28s is convened, the slaughter of Rooiland is ritually re-enacted. The overcrowded cells in which gang practice unfolds are transmogrified into the wide-open plains of the 19th-century highveld. The participants of gang practice are transported in space and time; through the metaphors that define their practices, they are Nongoloza, Kilikijan and their 13 followers, the original bandits of early industrial South Africa.

What are we to make of this, the fact that, in the course of practising their rituals, prison gangsters transport themselves from their jail cells into the mythical countryside of a bandit story?

The anthropologist, Mary Douglas, may be of some help here. She argues that "ritual focuses attention [on collective, existential problems] by framing [them] … The frame marks off the different kind of reality that is within it from that which is outside it… Ritual aids us in selecting experiences for concentrated attention."12

In other words, ritual and myth are to be understood as a form of collective problem-management. Those aspects of prison life that cause perpetual, ineluctable crises, that haunt the very existential beings of prisoners, are "framed off" from the rest of prison experience, placed in the myth and "worked on" through ritual. In Douglas's words, they are "selected for concentrated attention".

If this is correct, it follows that the content of the Nongoloza myth contains, in condensed and symbolic form, the central problems of prison life, and that gang ritual consists in an attempt to manage these problems.

So which problems have been placed in the myth and how does gang ritual deal with them? Perhaps the best way to begin answering these questions is to examine the Nongoloza story through its inconsistencies.

Consider the tale from Po's perspective. His story is a strange one—isn't it?—clearly invented in retrospect by men who already know that they are to spend much of their adult lives with other men, and away from their families. For the thinking of the old African seer is somewhat unfathomable. A man of his village and the embodiment of its interests, you would imagine that his task is not merely to save the young migrant workers for their own sakes. His initial concern, remember, was the village itself: the disappearance of those who were to secure its posterity.

Bandits who protect their peasant villages from extinction surely live alongside their communities, sheltering them from the forces that threaten to tear them apart. They are an adjunct to a way of life. But not in our story. Po's young men have no plans to return to their villages. They are a new breed, an eternal army that apes the one they fight, and they work for themselves, each other and the posterity of their band. Right at the beginning of their bandit lives, they have already forgotten their villages and have become embroiled in making the laws and mythology of their own cult-like future.

It would not be unfair to say, then, that the path Po chose was as corrosive of traditional life as the white men's gold mines. Just as the whites stole the villages' progeny, so Po stole it too, swallowing up the young migrants who leave their ancestral homes, never to return.

And how, it is fair to ask, are the outlaws to acquire their own progeny? There are no women and children. Who is to pick up the shield when Nongoloza dies? Surely, in the background to the pillages and murders that are the heartland of the tale, there will be future generations of migrants and drifters, young men who abandon their families and take to the hills. The ravages of colonialism must continue if the band is to survive. Their ancestral homes must be torn apart in perpetuity if the outlaws are to have a future.

And so there is an incoherence at the heart of the tale. The survival of colonialism is the band's primary nourishment. They have become bandits for the sake of being bandits. If the band is to live, so must injustice.

The incoherence of Po's actions, I will argue, stands in for or symbolises a central dilemma of inmate life in the prisons of 20th-century South Africa. The dilemma, in essence, is how to shore up one's dignity without destroying it in the very process of shoring it up. Put another way, Po's story is a bulwark against nihilism.

In making this argument, I begin with a general discussion on the conditions of imprisonment.

Several of the 20th century's finest theorists of inmate culture have spoken of the inherent degradations of life in total institutions.13 Inmate experience is in essence one of infantilisation. If normal adult life is made meaningful by the exercise of one's agency, this is precisely what is robbed from the inmate. Agency, here, refers not only to the overarching projects of adult life, such as raising children, developing a lifelong sexual partnership, forging a career, and so forth. Even the minutiae of adult agency—the simplest things we do by ourselves, under our own control, like wash, use the telephone, decide to eat, to rest—are taken away from the inmate. He is robbed of the very basics of the adult world.

Erving Goffman speaks of the inmate experience as one of "stripping" or "mortification", as the inmate is gradually robbed of the tools of his own agency. "Upon his entrance to the total institution," Goffman argues, "the inmate … begins a series of abasements, degradations, humiliations, and profanations of self. He is systematically, if often unintentionally, mortified."14 Indeed, Goffman continues bleakly, "Every total institution can be seen as a kind of dead sea in which little islands of vivid, encapturing activity appear".15

Similarly, Gresham Sykes, in his study of an American maximum-security prison, argues that:

[the] frustrations [of captivity] … carry a more profound hurt as a set of threats or attacks which are directed against the very foundations of the prisoner's being. The individual's picture of himself as a person of value—as a morally acceptable, adult male who can present some claim to merit in his material achievements and his inner strength—begins to waver and grow dim.16

Whether the mortifications of inmate life can be arrested, or even eliminated, by institutional reform has been the subject of much debate.17 For Goffman, though, the tendency of total institutions to erode the conditions of adult agency is ineluctable. No matter how much a total institution is reformed, it remains, by definition, a place in which control of the mechanics of daily life is transferred from adults to their custodians. Sykes, talking of the prison, and in particular of its primary function—to keep the inmates inside—makes a similar point:

Searching cells for contraband material; repeatedly counting all inmates to ensure that each man is in his appointed place; censoring mail for evidence of escape plans; inspecting bars, windows, gratings, and other possible escape routes—all are obvious precautions. The custodians, however, do not stop at these, for they have found to their bitter knowledge that in a maximum-security prison the most innocent activity may be the symptom of a major breach in the institution's defences. Pepper stolen from the mess hall may be used as a weapon, to be thrown into the eyes of a guard during a bid for freedom… A fresh coat of paint in a cell may be used by an industrious prisoner to cover up his handiwork when he has cut the bars and replaced the filings with putty.18

For both Goffman and Sykes, inmate culture is to be understood as a series of accommodations or adjustments to the degradations of inmate life. According to Sykes:

Frustrated not as an individual but as one of the many, the inmate finds two paths open. On the one hand, he can attempt to bind himself to his fellow captives with ties of mutual aid, loyalty, affection, and respect, firmly standing in opposition to the officials. On the other hand, he can enter into a war of all against all in which he seeks his own advantage without reference to the claims or needs of other prisoners.19

Goffman's thoughts on inmates' adaptations to the degradations of institutional life are more pessimistic, but they are also somewhat subtler. "Solidarizing tendencies … are limited," he avers. "Constraints which place inmates in a position to sympathize and communicate with each other do not necessarily lead to high group morale and solidarity."20 Rather, he argues, each inmate harnesses a more or less private response to the institution, one that aims to protect his sense of self against the violence of mortification. Yet even here, "the individual finds that his protective response to an assault upon self is collapsed into the situation; he cannot defend himself in the usual way by establishing distance between the mortifying situation and himself".21 In other words, the inmate's very response to his mortification is symptomatic of his mortification. It is, after all, an inmate's response.

With this in mind, Goffman identifies four categories of inmate adaptation to total institutions. The first he calls "situational withdrawal"—a mental flight from the institution. "The inmate withdraws apparent attention from everything except events immediately around his body." The second mode of adaptation he calls "colonization"—the inmate appropriates and valorises institutional life, rationalising that it is far better than life on the outside. The third he calls "conversion"—"the inmate takes over the staff view of himself and tries to act out the role of the perfect inmate". The fourth he calls "the intransigent line"—the inmate rejects the legitimacy of the institution and engages in a broadside against its authority.22

Spend any amount of time talking to Pollsmoor Prison's inmates about their incarceration and you will quickly identify all four modes of Goffman's adaptation. Spend an extended period and you will discover that many inmates deploy a combination of two or more modes of adaptation. As I was wandering through Pollsmoor's B section one day, a prisoner accosted me and began railing against the prison. He said Pollsmoor was too lax, that it didn't enforce the rules properly. "Look at how many inmates are walking around with earrings in their ears," he noted. Then he took out his copy of the Correctional Services discipline manual and showed me the clause stating that inmates are not permitted to wear jewellery. "At Helderberg Prison," he continued, "if the warders found you wearing an earring, they would rip it out your ear. I want to go to Helderberg. I will be at peace there."

The following week he had forgotten his annoyance at the laxity of rule enforcement in Pollsmoor. He was dreaming of the day he would be released (some time in 2007) and run a fleet of taxis in his hometown of Knysna. With a pencil and notepad he was calculating the differences in his daily profit if his average round trip from the township to the town centre was 50 per cent full, 60 per cent full or 70 per cent full. The following day he had pencil and notepaper out again, this time calculating what a 10c increase in the fuel price would do to his profits.

"You are talking like you are going to be released tomorrow," I commented.

"It will be like tomorrow," he replied. "For the next five years," he explained, tapping his temple with his forefinger, "I will be living in here."

But Pollsmoor is also home to the Number gangs, and they may well have surprised Goffman. They display a degree of oppositional militancy and an enduring solidarity that he dismissed as impossible—by fiat, it should be said, rather than argument. Nonetheless, it would be imprudent to discard Goffman's help in understanding the Number gangs. He may not have imagined their existence, but the tools he crafted can certainly assist in making sense of them. True, the gangs should never be conflated with the adjustments and adaptations of individuals—you will see later that many inmates' relation to their own gangs is acutely tactical and utilitarian. However, the gangs can certainly be understood as an institutional expression of a collective adaptation to mortification—an adaptation laden with all the irony and pathos Goffman so masterfully captured; indeed, they should be understood so.

In their most idealised self-representations, the Number gangs are a collective variant of Goffman's "intransigent line". They reject in toto the authority of the custodians and place a premium on inmate solidarity. Their representation of the prison is one of a war zone: two armies, each endowed with its own rank structure, uniforms and paraphernalia, are pitted eternally against one another.

The power of the Number gangs as a mode of adaptation to the mortifications of prison life lies in simplicity. What better way to shore up the agency prison has stolen than to borrow the agency of one's custodians? Instead of holding up an image of inmates to prison warders, the gangs hold up a mirror. "We are what you are. You are an army, we are an army. Where you have a head of prison, we will have a judge. Where you have a head of section, we will have a general. Whatever you do to us, we will do to you in turn."

There is, of course, no such thing as a purely "intransigent line". As you will see in the following two chapters, the relationship between gangs and warders has always been managed by a delicate filigree of unwritten rules. Both sides respect an invisible line, which divides permissible from unacceptable violence. But as an inmate ideology, the idea of forming oneself into a mirror of one's captors is a very powerful one.

Powerful, but costly. The price paid for aping one's custodians is very dear indeed. In a bitter irony, prisoners end up re-establishing and practising upon one another the very system of mortification the custodians unleash on them. Consider, for instance, the nomenclature of Number gang punishment, and you will find that much of it is borrowed wholesale from the nomenclature of institutional punishment.

One of the more serious forms of punishment the gangs mete out to offending members is called the "carry-on". The offender stands with his arms in the air, leaving his torso exposed. The punishers form a circle around him. One of the punishers cries "Up!" and the rest batter the offender's torso with padlocks and bars of soap, each placed in a sock. This is a simple imitation of an informal punishment warders meted out to gangsters in some prisons during the apartheid era. When, for instance, a 26 stabbed a warder, the prison's custodians would arm themselves with batons, baseball bats and knobkerries. They would assemble every 26 member in the yard and form a circle around them. A warder would cry "Up!" and the rest would attack the assembled 26s with their batons and knobkerries. This was called a "carry-on".23 The Number gangs have simply appropriated it, and deploy it to punish their own members.

More disturbing, perhaps, is the archetypal relationship between Number gangsters, who are called ndotas (men), and non-gangsters, who are called franse. Like the passive partner in a typical sexual relationship behind bars, the frans has been stripped of the jail equivalent of his juridical personhood. When he receives a parcel from a visitor, he must hand it over to the ndotas in his cell; they will decide how it is to be distributed. If he wants to conduct a commercial transaction—sell his watch, swap a T-shirt for a toothbrush—he must ask the permission of the ndotas. In exchange for allowing him to conduct a transaction he must give the ndotas something in turn. When ndotas in a cell hold a meeting to discuss Number business, each frans must sit with his face to the wall and remain absolutely silent. Franse are also servants. They keep the cell clean and wash the ndotas' clothes.

One way of understanding the relationship between ndotas and franse is as a strange throwback to the feudal realm: a frans rents the very air he breathes. Another way, of course, is to understand the relationship as an enactment of the very mortification Goffman describes. Like Goffman's inmate, the frans is robbed of the minutiae of adult agency. He is, quite literally, told when to shit and when to eat, when to stand and when to talk. He has no right to personal possessions, to any of the material accoutrements around which a person forms his own quiddity and individuality.

One is reminded of Goffman's comment cited earlier: the inmate "finds that his protective response to an assault upon self is collapsed into the situation; he cannot defend himself in the usual way by establishing distance between the mortifying situation and himself". Indeed, in attempting to escape mortification, the Number gangs become quasi-institutions of mortification themselves. Having pushed the custodians to the periphery, and having carved out a space of ostensible autonomy, they use that space to imitate and re-enact the very precepts of the total institution. A cynic might argue that all the gangs have managed is to double the total institution over itself, creating a system of mortification within a system of mortification. The gangs' imagination is exhausted by the very universe against which they rail. The relationship between custodian and inmate is ubiquitous; all the gangs are able to do is to repeat it eternally.24

In this sense, the Number gangs are quintessentially reactionary institutions. They are locked in a circle, repeating the conditions of the present over and over.

This argument is, of course, a little too stark and needs to be qualified. Gang practice does foster real masculine virtues such as bravery, stoicism and solidarity. These are precisely some of the precepts of adult agency that the process of mortification strips, and there can be little doubt that the gangs have managed to retrieve them for at least some of their members. The point is that the cost of this retrieval is very severe: the gangs are only able to shore up some of the precepts of agency—in this case, certain masculine virtues—at the price of re-enacting the custodian–inmate relationship.

There is a related but subtly different way in which the Number gangs are reactionary. They are nourished and animated by the violence of the custodians and the degradations of the prison. In ways that will become clearer in the following chapter, the virtues that the gangs celebrate—controlled and meditated violence in the face of provocation, stoicism in the face of brutality, solidarity in the face of mutual adversity—are all fostered by the severe conditions of the apartheid prison. Later, you will see that when the prison regime softens, the Number gangs tend to splinter, and sometimes even to disintegrate. In this sense, the gangs need the very degradations against which they rail. It is truly astounding how many of the veteran gangsters I have interviewed speak nostalgically of apartheid, of "the old days" when, as one interviewee put it, "the violence of the boere was predictable and kept the Number strong".

The gangs are thus locked in an eternal relationship with their captors. They have no project, no goal, no horizon towards which they move.

The essentially reactionary nature of the Number gangs—their reliance on and replication of the violence of their captors—is a source of deep discomfort among Number members. Indeed, the idea that they have no goal, no project, is intolerable. Without exception, every veteran ndota I interviewed described the Number of the apartheid era as a revolutionary, anti-apartheid army. The following comment, by an ndota who spent much of the 1980s and early 1990s behind bars, is typical:

"Okay, I grant you, we couldn't smash the whole apartheid state because we were locked in jail. But we did our best to destroy that part of the state we had access to: the prisons. Our task was to make them ungovernable, and eventually to break them down."

"How did you plan to do this?" I asked.

"Well," he replied, "not quite break them down, but improve them. We were fighting for our rights. It is thanks to us, to the thousands of us who were killed and tortured and given carry-ons, and locked in solitary confinement without proper food, it is thanks to our broken limbs and our blood that we won the rights we have today."

"What sort of rights?" I asked.

"You should have seen this place 15 years ago," he replied. "We had no beds, just two thin blankets, so thin that if you put them together and held them up to the sun, your eyes would still get burnt. We were not allowed to wear watches. We did not have televisions. We won each of these rights with blood and sacrifice."

There is a great deal of pathos in these comments. The idea that a century-old anti-colonial army, painstakingly preserved and transmitted across the generations, adorned with imaginary uniforms, a complex, finely observed military structure, a rich myth of origin—the idea that such an entity exists to fight for a bed, a thick blanket, a watch is a humiliating one.

It is even more humiliating when one considers that the story is empirically false. It is almost certain that the small improvements in prison conditions which began in the late 1980s—beds, televisions, watches—were the result, not of Number gang activity, but of the campaigns organised by and in support of political prisoners. It was political and not criminal prisoners who began using the courts to contest the conditions of their imprisonment. It was the prison memoirs of prominent political prisoners that finally brought to the world testimony about the conditions in South African jails.25 And it was the imprisonment of global icons such as Nelson Mandela that brought global attention to penal practice in South Africa.

Indeed, the relationship between Number veterans and the South African liberation movements is a telling one. I have interviewed several political prisoners who spent time in the criminals sections of South African prisons, and thus in cells run by the Number gangs, during the 1970s and 1980s. Some were castigated by Number gangsters as "communists", "agitators" and "troublemakers". In other words, the Number had swallowed wholesale the bluntest elements of apartheid's anti-liberation movement propaganda. Others, however, tell a more complex story. According to Vincent Shabangu, a student activist convicted of public violence in 1985, who did most of his time in Pollsmoor:

Initially I was treated as an odd case. They weren't sure what to do with me. I wasn't an ordinary frans, but I wasn't a gangster. Then I started engaging with them about prison conditions, about prisoner rights, about what these things had to do with apartheid. Soon, they managed to categorise me: I was a special frans. While other franse had to get permission to leave their cells, I could walk around the section freely. When a newspaper was smuggled into the cell, it was given to me to read. They were begrudgingly in awe of my political knowledge. They used me as an interpreter of political life on the outside. At one point they told me I was so useful that I must join the 26s. I said I'd join, but only if I didn't have to go through all the shit they go through to join. I wanted to head their legal section immediately, their courts, because their legal system was crazy and I wanted to fix it. They said no, you can't join like that. So I remained a special frans. We watched each other's moves like chess players: mutual suspicion, mutual respect.26

Several MK (Umkhonto we Sizwe, the military arm of the African National Congress) soldiers who spent time in the criminal sections of South African jails say that ndotas regarded them with a mixture of circumspection and awe. Jeremy Veary, an MK soldier who was incarcerated at Pollsmoor in 1987, says this:

The boere threw us in the criminal section because they thought the gangsters would kill us. Instead, the gangsters checked us out, asked us lots of questions. When they found out we were MK, they went into a huddle and discussed what to do with us. I got the sense that it could have gone either way; they could have attacked us, or accepted us. Finally, they made a decision: we were to be looked after by the 27s. MK was the ANC's army, and the 27s was the Number's army. We were kind of equivalent.27

Grudging respect, but suspicion of Vincent Shabangu; thinly concealed idealisation of MK soldiers: the relationship of the gangs to the liberation movement was a fraught one. The liberation movement embodied precisely what the gangs aspired to embody: a goal, a project, to destroy apartheid. For this reason, they identified with the liberation movement on the one hand, but envied and wished to harm it on the other. For the liberation movement represented an actual embodiment of the very ideal that would always remain elusive to the gangs: the ideal of transcending the reactionary cycles of the relationship between prison gangsters and their custodians.28

And indeed, at the very moment the ANC came to power, in 1994, South Africa experienced the largest and deadliest wave of jail disturbances in its history. Between March and June 1994 (the elections were held on 27 April), there was unrest at 53 prisons across the country. Together, these prisons housed more than three-quarters of the country's jail population. In every case bar six, inmates attempted to burn the prison down. By the time the rebellion was subdued at the end of June, 37 prisoners had been killed and 750 hurt. No warders died, but 145 were injured.

The reasons for the disturbances were complex and manifold, but one of them was that the new ANC government refused to celebrate the inauguration of democracy by granting prisoners a generous amnesty. Veteran Number gangsters still speak of this with great bitterness. They describe their anger in the language of filial treachery; the ANC is guilty of a Cain-and-Abel betrayal.

I argued earlier that, when examined from Po's perspective, the Nongoloza myth operates as a bulwark against nihilism. It is the expression of a desire to transcend the horribly reactionary character of the gangs' adaptations to the mortification attendant on inmate life; an attempt to give the gangs an imaginary political project. Note how politically ambitious Po's story is. He does not simply valorise law-breaking under colonialism as a political good in and of itself. He attempts to position law-breaking as the centrepiece in a project of anti-colonial preservation. He is the embodiment of black interests; his task, initially, is to protect the posterity of his people in the face of colonial violence. That is why he trains young men to be criminals: to defend his people's progeny from annihilation.

The politics embodied in Po's story is possessed of an almost desperate ambitiousness. Crime is the substance of the liberation project. There is no space in the story for a rival liberation project. No wonder ndotas greeted the activists they met behind bars with such a fraught mixture of envy, anger and admiration. Activists were a mirror to the ndotas' impossible fantasy—their fantasy of cutting through the shackles of their Goffmanesque adaptations.

There is indeed much sadness in the speed with which Po's initial intentions unravel. Before he has properly formulated his anti-colonial politics, he is already embroiled in an eternally symbiotic relationship with his colonial masters. Understood as the metaphorical representation of a dilemma of prison life, Po's story is an expression of humiliation—the humiliation inherent in the inmate's strategies of adaptation.29

There is, of course, a range of other vantage points from which to interpret the Nongoloza myth. The most important of these, perhaps, is that of Magubane. The narrative did, after all, start as a tale of anti-colonial banditry, only to end as an irresolvable conflict over male sexuality. Magubane, as we shall see in Chapter 3, is the bogeyman of the tale—the greater tormentor of the Number gangs.

For the moment, though, I wish to turn to a subject raised briefly in this chapter, and one which ndotas are far happier discussing: the function of violence in the fostering of masculine virtues.

Chapter 2: The functions of violence I—Making men (and not children)

In early 2003, a Pollsmoor inmate by the name of Magadien described to me his initiation into the 28s. At the time of our interview, his memories were 25 years old; he had been initiated at Victor Verster Prison in November 1977. He was 18 then. The following is an edited version of his testimony, together with some commentary.

"How did your recruitment begin?" I ask Magadien.

He gives a strange answer, one that, I suspect, is drawn half from myth, half from his actual experience. It is said that when a new member is recruited into the 28s, the glas (a particularly colloquial expression that means "binoculars")—a senior member, one of whose tasks is to conduct gang business in "the bush", those parts of the prison where the 28s are not active—comes to the potential recruit and confronts him with a riddle. How the newcomer responds determines whether he will be recruited and, if he is, into which department of the gang.

"The glas watched me for a long time, the way I interacted with other franse—the way I dealt with conflict, the way I solved problems. He approached me and we talked, and after a while I said I wanted to join the 28s. During one of our discussions, he said to me: 'I'm going to ask you a question. Think carefully before you answer. It is raining. You are standing under an umbrella. I say to you, I am getting wet. I may get sick. What are you going to do?'"

"I answered correctly. I said: 'I will come out into the rain with you,' which means: 'I am prepared to live like you. We are brothers. We will live and die together.'"

"And if you had invited him to share your umbrella?" I ask.

"That means I am inviting him into my bed. If I had given that answer, they would not have made me a 28. They would have made me a sex son.

"A sex son is mixed with the probationers in the silver line of the 28s and, to an outsider, he is just like a probationer, a new 28. But actually, he is a sex object: the soldiers, the members of the gold line, sleep with him at night. He cannot progress up the silver line. He is never told much about the history. He is not really a 28.

"So I answered the question correctly, and was told that to join the gang I must stab a white warder. I was told when the knife would arrive, how I must do it, and what I must do afterwards.

"I actually thought the whole thing was a joke. We were thoroughly searched every day. I thought there was no way the 28s were going to get a knife to me. But they did it. I couldn't believe it. They said it would be Friday afternoon and it was. I picked up my tray of food outside the kitchen, and it was there, hidden away, this little knife, all wrapped up with bandages, with just the tip of the blade sticking out. I learned later that the nyangi, the doctor, had wrapped the knife. That is one of his main functions, to determine the length of the blade.

"I was told I must stab as soon as I saw the knife. So I brushed past a white warder, and when I was just past him, and his back was to me, I grabbed the knife, dropped the tray and stabbed him in the back, beneath the shoulder. They had told me it must be beneath the shoulder. I must not go near an organ or the spine—just a flesh wound.

"Then everything happened at once. All the inmates around me shouted 'Nangampela!30 Die nommer is vol!' (The Number is complete!) I don't know how the 28s got there in time. It was so quick. But suddenly there was a scuffle around me, I felt the knife being taken from my hand, and in a second it was gone. It was just me standing there in shock, with empty hands, and the warder standing there in shock. The knife was gone.

"The boere beat me, brother. They beat the shit out of me. By the time they had locked me in the one-ones, or agter die berge31 —in a single cell—I had pimples all over my body. I was so swollen I looked like a bodybuilder. But I did not cry out, not once. You see, the ndota who recruited me, the glas of the 28s at that prison, he told me: 'When you stab, the warders will beat you, there and then, in front of all the inmates, all the way to the one-ones. If you cry out, or moan, just once, your stabbing means nothing. It is as if you have stabbed nobody; you will never be a 28.' Then he showed me how not to cry."

Magadien takes a roll of toilet paper from next to his bed, breaks off a wad, and shoves it in his mouth.

"You do that," he says, "and you don't cry. You just squeak like a mouse, softly, so that nobody can hear you.

"After a couple of days in the one-ones, I noticed things started happening. The cells around me were being filled with 28s. They went there deliberately to be with me. They would break rules—like shouting, or pretending to fight—so that they could come to the one-ones. And they started talking to me. At exercise time—all the one-ones exercised together—they began probing me.

"I didn't realise it at the time, but what they were doing was putting steel in my blood, giving me support to prepare for what was going to happen next. Because what happens when you stab a warder is very rough. If you're a new recruit, you need to know that your brothers are waiting for you on the other side.

"After some time in the one-ones, the authorities took me to what we called the X court. The magistrate comes to the prison. There was a quick trial; it lasted about an hour. Two years were added to my sentence.

"Then they took me back to the one-ones. The boere said: 'Ja, hotnot. That was the magistrate's justice. Now for our justice.' They kept me in the one-ones, in segregation, 15 days of spare diet. They can't give you a saltless diet for longer than that. You will die. They had nutritionists who worked out how long you could go without salt."

After 15 days, when his internal sentence was over, Magadien was taken to the maximum-security section of the prison and put in a 28 "tight corner": a cell dominated by the 28s.

South Africa's prison gangs are perhaps the most ritualised structures you will ever find. Three days of the week are reserved for a carefully delineated set of functions. Friday is the day of the rations. It is then that all the possessions in the cell are gathered, including the possessions of franse, and are distributed to individuals: first, according to whether they are franse or ndotas, and then among ndotas, according to their rank. Friday's task is performed within the confines of the cell.

But Saturdays and Sundays are different. On these days, the central structures of the gangs in each section of each prison must meet, and they do, and warders have tacitly allowed them to do so, as I discuss in the following chapter. Saturday is the day of the wrongs, Sunday the day of the rights. (The gangs say "the year of the wrongs". Everything about the metaphors of prison gang language is expansive. An overcrowded cell becomes a vast plain; a day becomes a year.) On Saturdays, the various judicial structures of the gang meet to pass sentence: an errant member is punished; a new recruit is scanned for illicit allegiances. On Sundays, newcomers are recruited, members are promoted, victories are celebrated.

On his first Saturday in the criminal section, Magadien was placed in the centre of a cell and told to strip down to his underwear. The glas circled him slowly, scrutinising his skin for vuil papiere (contaminated papers)—the tattoos of other gangs. If the glas had found the mark of the 26s or 27s on Magadien's body, negotiations with the rival gang would have begun. If he had found the mark of a prison gang the Number consider illicit—like the Big Five, whose aim is to spy on the Number on behalf of the warders, or the Air Force, whose aim is to escape from prison—Magadien would have been beaten and his recruitment halted.

On the Sunday—the year of the rights—his recruitment ceremony is performed. He is again placed in the centre of a cell, but this time several people surround him. The first to approach him is the nyangi, the doctor. What happens next is metaphorical, and is enacted by a simple process of physical mimicry. On the shoulders of the nyangi's imaginary uniform are twelve pipes: six are gold, six silver. He tells Magadien to hold out his arms, palms upwards. He takes a gold pipe off his shoulder and slaps it on Magadien's right wrist, then a silver pipe on his left wrist. He checks Magadien's pulse and declares: "Die man se pols klop twee keer per jaar" ("The man's pulse beats twice a year"), which means that Magadien is being recruited into the silver line. If he had said three times a year, it would have meant the gold line. Then the gwenza—a senior member of the silver line—places a handkerchief on the floor and slips a knife under it. He stands up and addresses Magadien:

"From today," he says, "you are no longer a frans. You are a 28. You will never swear at your brother. You will never hurt your brother. You will never do anything that reflects badly on the camp. And if you leave the camp, you will leave by your own blood.

"I am giving you your uniform. You have a white pair of sandals. Your socks are also white, stamped with the sign of 28, both inside and out. You have a white shirt, and white belt with silver buckles. You have a white tunic with two buttons, stamped with the sign of 28, inside and out. You wear a green tie. You have a white jacket, which also has two buttons, the first open, because you belong to the Number day and night, the second closed, for discipline. You wear a white beret with a silver badge, and on the badge a hammer and a handkerchief are engraved."

Then the landdros (magistrate) comes forward, takes out his green and white stamps and gives Magadien's recruitment his approval. The landdros carries four stamps—white, green, red and black—signifying the four hooves of Rooiland. When a silver is recruited or promoted, the landdros takes out his green and white stamps. When it is a gold, he takes out the green and the red, the red signifying blood. His black stamp is reserved for the death sentence.

Next, the magistrate steps forward and takes out his imaginary white and green pens to inscribe Magadien's recruitment in the 28s' proverbial record book. Like the landdros with his stamps, the magistrate has four pens, which signify Rooiland's legs.

Now, Magadien is marched out of the circle, and the formal recruitment ceremony is over. The 28s take the news of his recruitment to the 26s and 27s. They do so in the form of an exacting ritual.

Every night, the glas and draad of the 28s meet with the glas and draad of the 27s, and the glas and draad of the 26s, in a forum called the "Valcross". Only the glas is allowed to speak on the Valcross. The draad must remain silent; he is the one who will report back to the 28s what happened on the Valcross. The two 28s on the Valcross are not allowed to speak directly to the 26s. They communicate through the 27s. (This is an enactment of the principle that the 27s are the ones who uphold gang law, protecting the 26s and 28s from one another, and mediating their relationship.) At the Valcross meeting on the evening after Magadien's recruitment, the glas "bugles" his recruitment to the 27s, who, in turn, pass on the information to the 26s.

In the week to follow, Magadien will sleep alongside a different member of the 28s every night, and each will describe his uniform and his functions. The structure he himself is to join, once his training is over, is called the four points of the twos ("two" is a reference to their pulse—it beats "twice a year"). It is the lower tier of the silver line. He learns that one of its functions is that of a court: it deals with infractions committed by low-ranking silvers.

The five members of the twos are, from most junior to most senior, the silver-two, the silver-one, the Goliat-two, the Goliat-one and the landdros. The silver-two is responsible for the security of every meeting of the twos. When a junior member is to stand trial for a minor infraction, it is the silver-two who searches him for weapons, leads him into the circle, and then leaves the circle again in order to guard it. The silver-two is also responsible for the sex sons: if a soldier forces a sex son to have penetrative sex, he must report the incident to the silver-two.

The silver-one is the intelligence officer of the twos. He tells the court what the culprit is accused of, and the circumstances of the case. The Goliat-two is the keeper of the laws of the twos. He will determine the punishment at the conclusion of the case. The Goliat-one is the defence lawyer in the silvers' misdemeanours court: he will argue the accused's case. And, finally, the landdros, the most senior member of the twos, is the prosecutor. He also "has one foot in the ones", the upper tier of the silver line, which Magadien will join in the mid-1980s: he is the "centre post" between the upper and the lower tier of the silvers.

It is not from the four points of the twos that Magadien learns their functions, but from senior members of the gold line whom he sleeps next to every night. They too have their own "sub- court" to deal with the infractions of gold-line members. Their court mirrors precisely the functions of the lower silver-line court. So they explain to Magadien their own functions, and in so doing teach him the functions of the silver-line court he will join. They also begin to teach him the history of Nongoloza and Kilikijan. But only a fragment of the history—he is still too junior to be told the whole story.

Finally, he is taken to a section of the gang called the mambozas, or the forties. The 28s here are senior but inactive. They are either too old or too injured for active duty, or their position in the structure is already filled by someone else. For instance, the 28s cannot have two active nyangis in one prison. If a second nyangi is transferred to Victor Verster, he is dormant—he sleeps in the forties.

The mambozas begin to teach Magadien to sabela—to speak prison language—and it is a long, gruelling process. "Your blackboard [teacher]," Magadien tells me, "sabelas with you day and night. Your first two months in the Number you are not allowed to receive visits, to read or write letters, or to read books. You must focus on the Number, and when you learn too slowly the punishments are severe. If you cannot remember something you have been taught, you are stripped and stood under a cold shower until you 'find the Number'—until you get it right."

One of the last things Magadien learns about the structure of the 28s takes the form of a ghost story. There is a position in the silver line held by a man called Mtjoetjies. But he is dead; his place in the hierarchy is left empty.

"How did he die?" I ask.

"Back in the beginning," Magadien tells me, "when Nongoloza was imprisoned at The Point, he refused to learn English or Afrikaans, or any language spoken by white men. He refused to look white men in the eye. But one of the 28s, Mtjoetjies, was fluent in English, and Nongoloza used him to negotiate with the boere. After a while, Nongoloza realised that he had made a mistake. 'What does this man talk to the boere about? I cannot understand what he says. He may be betraying us.' So Nongoloza killed him, as a precaution.

"And now, his position remains there, empty, as a reminder that we do not negotiate with words, we negotiate with action, with violence. There have always been social workers in prison, you know. And they want to talk to you behind a closed door. If you do that, if you talk to an authority alone, one-to-one, you are severely punished."

Magadien's account of his recruitment is undoubtedly an idealisation. There is something unsettlingly ethereal about his recollections. The world he describes is entirely unpeopled: there are nyangis, mtshalis, gwenzas and silver-twos, but there are no flesh-and-blood beings. It is as if he has dissolved the concrete and the bars, the bad food, the bile that passes between men confined together, and abstracted from it a pure form. The gang he speaks of could be lifted and taken to any prison, anywhere. It is not about this jail, that cell, those people—it is just an idea.

Nonetheless, as an idea, there is much about it for us to explore. Magadien's is a rich, layered account and there are many aspects of it I will return to later. For the moment, I wish to explore one theme in particular: not the violence that Magadien must commit in order to join, but the violence to which he must be subjected.

It is not enough for Magadien to stab a white warder. In and of itself, this means very little. He must be beaten heavily after the stabbing, publicly, in front of the other prisoners, and he must not cry out during his beating. He must be taken agter die berge—to solitary confinement—and endure a time of isolation and spare diet. And yet "endure" is probably not the appropriate word here. It is not enough that he survives his isolation, that he just pulls through. He must not emerge shaken or bewildered; he must emerge "with steel in his blood".

Ndota means "man" in Zulu, and there can be little doubt that the recruitment ritual is an initiation into manhood. But what sort of man? The being created by the initiation is, really, a distilled version of a man—the sort of man a second-rate philosopher cobbles together. He consists of a few hard, shorn virtues and nothing else. One is solidarity, but solidarity of a very particular type: the recruit is robbed of a part of his own physical integrity for the sake of the gang. His sacrifice is visceral and highly personal; he feels it on his body. A second is stoicism. It is critical that the recruit shows indifference to his pain. He must not cry out when beaten. He must emerge from isolation unscathed. This is, perhaps, the starkest and most striking of the virtues the initiation fosters: the recruit must demonstrate the ability to amputate, or at very least to stunt, vital needs, feelings and responses. A third, paradoxically, is measured restraint: paradoxical because restraint is exercised in the very process of stabbing. The stabber can only use a short blade, one designed to inflict a surface wound. He must target a safe section of the body, one where he knows he will not cause serious or permanent injury. You will see later that this is among the most sacred of the Number's virtues, and that the survival of the gang hinges on it. A gang member who cannot restrain himself, who seriously injures a warder when he was not mandated (or "dutied" in gang language) to do so, is severely punished.

It is helpful to return to Erving Goffman to understand the initiation ritual, for the virtues fostered therein are, surely, the fruit of a particular mode of adjustment to mortification. If the prison's tendency is to infantilise inmates, the ritual's aim is to make inmates into men. They are, as I said earlier, not whole men, but distilled, shorn aspects of whole men—inmate men. They are little more than a collection of stark, honed virtues. But there is little doubt that the newly recruited ndota has retrieved, in a partial and somewhat damaged form, some of the attributes of adult agency. Of course, the agency the recruit has retrieved will be used to mortify in turn. He will use his power to inflict the institution's punishments on his own brothers, to turn franse into exaggerated versions of the inmate.

If the initiation ritual is indeed a form of adjustment, a retrieval of agency, note how it deploys the very tools of warder violence to do its work. The initiate must be beaten publicly. He must endure a spare diet in solitary confinement. The gangs have appropriated punishment and transformed it into a vital tool in the passage to manhood. In a strange perversion, they need the violence that is meted out against them. They have transformed that violence from a tool of mortification into a form of nourishment.

Note also that the initiation ritual has carefully sifted through the various forms of institutional punishment, taking those elements that it needs. In Magadien's case, institutional punishment was three-tiered. There was the formal punishment of the criminal justice system: a magistrate came to Victor Verster and a criminal trial was conducted. Magadien had time added to his sentence. For the Number gangs, the function of this punishment was for Magadien to demonstrate the dissolution of his individual interests in the face of the interests of the gang. He was choosing to delay his own freedom in order to rise through the ranks of the 28s. He was, quite literally, suspending his own life in order to step into another universe, an all-encompassing brotherhood.

But the meat of the ritual is found in the gangs' appropriation of the second and third tiers of punishment. The second was the formal punishment the warders handed out: isolation with spare diet. Finally, there was the warders' informal punishment, an illegal punishment, written, not into legislation, but into the informal annals of warder practice: he was heavily beaten on his way to isolation.

There is a great deal of irony at work here. The informal mode of punishment presumably arose because warders believed that the punishment available on the statute books was not adequate to control the prisons. Of the dozen or so white warders I interviewed, not one of them appeared to have contemplated the possibility that their informal and illegal violence was, in fact, a source of vital nourishment for the gangs.

If the initiation fosters virtues of solidarity, stoicism and restraint, to what end? To what work do the gangs put these virtues? Consider the following testimony. Once again, the speaker is Magadien. He is talking of a stabbing that took place in 1982. This time, the stabber is not Magadien himself but a fellow 28 whose gang name was "Cups". The prison is not Pollsmoor but one of the small rural jails that dot the Western Cape countryside. They were, in essence, agricultural labour stations. The farmers of the district used the inmates to work in their fields in exchange for various forms of payment.32

"There was this warder called Malan," Magadien begins. "Jesus, he was verkramp. He would come up behind you and smack you on the ears, for nothing, no provocation. Malan and Cups had a quarrel. I don't remember what it was about. I think it was about clothes. Malan was responsible for issuing clothes. He was stingy. Some of us were walking around with holes in our trousers so big it was like not wearing trousers at all. It was undignified. Cups went to him to complain about the state of our clothes, and he treated Cups like shit, in front of the other ndotas, told him to go fuck himself, or something like that. Later, the glas of the 28s, the one whose job is to observe and gather information, came to Cups and said: 'Why does that mapuza start with you like that all the time?'

"He didn't have to say what he really meant; it was obvious. He was saying four things: one, you are losing your dignity; two, the Number needs to make this man afraid; three, if you stab him, you will be doing a duty and will get promoted; four, if you don't deal with him, we will punish you, because to allow that man to get away with what he is doing puts the whole camp at risk.

"You see, there are good warders and there are bad warders. A good warder, we say about him that he keeps his office clean. He will order us around, but only to do things that we respect, like sweeping and cleaning. That is good. Everyone needs a clean prison, even the Number. He will respect us as human beings: he won't deny people meals; he will listen to complaints; he will tend to a sick inmate; he will not openly say that the Number can operate, but he will respect the dignity of the Number.

"But Malan was a bad warder: he is one who thinks he can get away with treating an ndota like shit, and that is one thing he cannot do. Because if you allow that then, slowly but surely, the Number can no longer operate in that prison. What would happen if we could no longer meet on the year of the wrongs? We can't allow that to happen.

"So, the order came down to Cups: 'Skiet hom, so dat hy jou respek.' ('Shoot him, so that he respects you.')

"Cups was given a knife that had been inspected by the nyangi, and when he went to stab, he was accompanied by the glas and the draad. Their function depends on whether the stabber does his job. If he stabs, they make the knife disappear. If he chickens out, they do his job for him and they punish him—they stab the warder and the ndota. So Cups walked up to him and stabbed him just like that. He was talking to another inmate; his back was turned to Cups. The glas and draad were with him and the knife disappeared.

"They beat the shit out of Cups and put him agter die berge. But they did something else as well. Although they could not find the knife, they knew it was Cups. And because they knew it was Cups, they knew it was the 28s. They know that nobody stabs without being dutied to do so.

"They called the local farmers' association over the two-way radio to tell them that a white man had been stabbed. And all the farmers from the district arrived in their jeeps and bakkies. They came with pickaxes, knobkerries, sticks—anything they could find. Then, the warders assembled every single inmate in that prison in the courtyard. They surrounded us and then the warders separated us. "Franse here, 26s there, 28s there. Then they forced the franse to take off their clothes so they could search for tattoos. If they see that a 28 member has pretended to be a frans, they put him in the right group. Once that is done, they put the franse in the cells. Then the 26s. It is just the 28s left in the courtyard, and the farmers are all around us with their weapons. A warder shouts: 'Up!' And all hell breaks loose. They break bones: kneecaps, legs, collarbones. Because they know it was the whole camp that made the decision to stab, so the whole camp must suffer."

"Didn't you fight back?" I ask.

"No, the rule was that you never provoke. If you provoke, if you fight back in a situation like that, they will kill you. It was all tactical. Sometimes it was necessary to retreat, for the sake of the survival of the Number.

"Sometimes, it wouldn't happen in the yard. Sometimes it was in the showers. They would wet the floor of the showers, so it was slippery, and we would have to take off our clothes. You can't run. You are barefoot, and the ground is slippery. They are wearing rubber soles. Then they beat the shit out of you, and you don't fight back, because if you do, someone will die."

What struck me, as I listened to Magadien, was not so much the brutality of the relationship between warders and inmates as the intricate filigree of unwritten rules and corresponding tactics. Inmate violence against warders in this particular instance, it seems to me, served a specific function. It punishes a warder for breaking two unwritten rules: first, that warders treat ndotas as ndotas, not franse, that they respect the dignity an inmate has earned through his recruitment into manhood; second, that warders respect the circumscribed spheres of autonomy that the Number gang have won: to meet on the year of the wrongs and the year of the rights, for instance.

There is thus a circular logic to this aspect of gang violence. One commits and subjects oneself to predicable ritual violence in order to forge a few simple virtues. And then one triggers the same ritual of violence in order to defend the space in which one practises those very virtues. This function of the violence is thus self-reinforcing: violence protects the very virtues it creates.

Magadien's "good warder" is one who, unlike Malan, has read and understood the unwritten rules and obeys them. He knows that for the sake of both his own self-preservation and for the sake of order in the prison, he must tacitly allow the Number gangs a certain amount of space; not too much space, because that too will threaten order. It is a question of intuition that comes with experience.

"It was a bit like a game," Magadien says. "The question is: who's afraid of whom at the end of the day? We chose warders whom we needed to fear us, and we stabbed them. Sure, they fuck us up afterwards, but put yourself in the shoes of the individual warder, the one who has been stabbed. You go away for a while on sick leave, maybe a little longer on stress leave, but then you are back. When you walk in the yard, with all the prisoners around you, you are worried. When you walk alone down the corridor, or into a cell, your heart is beating fast. Maybe, in the front of your mind, you think these people stabbed you just because they are animals. But at the back of your mind you know why you were stabbed. You do not shout, you do not scream, you do not clip a prisoner on the back of the ear. You are afraid.

"Why do you think the Number was allowed to meet three times a week: on the year of the rations on Fridays, the year of the wrongs on Saturdays, and the year of the rights on Sundays? Why do you think the warders never broke those meetings up? Because they knew. Maybe they didn't know the details. They didn't know that J.R. is a silver-two and his job is to stab me if I interrupt that bunch of inmates. But they sensed the atmosphere. They knew that, for their own safety, there are places they must not go."

Inmates, too, knew that there was a threshold they must not cross. When an inmate was dutied to stab, it was a sacred imperative that the nyangi determine the length of the blade. It ought never be long enough to kill. And the stabber was also issued with a strict instruction to strike the flesh on either side of the spine, never a vital organ. When the gang was collectively punished for a stabbing, their own self-imposed rule stipulated that they do not fight back. They knew that injuring a warder critically or fatally, or resisting ritualised punishment, would invite a deadly response, and they avoided it like the plague.

On the few occasions that an ndota did die, the gang suffered acute trauma. When a gang member was convicted of murder, he was usually hanged. On the day of the hanging, every member of the dead man's gang in the prison where he served shaved his head. For the following month, they walked the passages of the prison in absolute silence, their heads bowed and their hands crossed at their waists, a symbol of inaction. Members only spoke with one another in their cells, after lockup, and only in hushed tones. And they banned all gang activity: promotions, punishments, all decisions. The members of the other two gangs knew that the gang in mourning would not be sending representatives to the Valcross; they observed the abstention with due respect. Only once the month was up, and the stubble had begun to appear on their heads, did they begin to speak again, and to thaw the rituals they had frozen.

They took care to avoid death, to be sure, but the threshold of violence they were prepared to tolerate was very high indeed: broken kneecaps and collarbones. And, in the aftermath of a ritual beating, they did not receive due medical attention.

"We never trusted prison doctors," Magadien tells me. "For one, they were white. And they were also mapuza, the enemy. We called them perdedokters (horse doctors). There were a lot of inmates who stumbled round the prisons with broken bones, a lot of people who were stuck in the mambozas the rest of their lives because they were too fucked up to be active."

Was it worth it? How did the damaged limbs weigh up against the value of winning the space to conduct gang business? In the following chapter I will give an account of gang breakdown, a war of all against all that broke out across the prisons in the Western Cape in the late 1980s. What happened, it seems, is that the soldier lines of the gangs cracked under the violence that was meted out against them. They turned on the non-combatants in the gangs, and when the fight was over, they tore up the gangs' most sacred rules.

The notion that order in the prisons was maintained by a delicate system of unwritten rules is one that begs for a control experiment. What happens when one side breaks the rules, when either ndotas or warders threaten to kill one another? Interestingly enough, there was a kind of control experiment in the 1980s: the medium-security, criminal prison on Robben Island.

"At Robben Island," a veteran inmate told me, and his testimony was echoed by many others, "there was one phrase: 'Everything dies in the sea.' Nothing that happens reaches Cape Town. You know that if you stab a white warder at that place you die. Who wants to die like that? We knew that. The warders knew that. That's why they sent us there. There were lots of 28s and 26s there who had once caused trouble at Pollsmoor.

"But it became a nice place if you behaved, and you did behave because you didn't want to die. The warders there were very corrupt. They'd choose two or three inmates, give us wetsuits and take us into the water to dive for crayfish and abalone. Then they'd sell it on the side to line their own pockets. And you know what? When they took us diving they gave us knives. Each inmate would get his own knife, hardened gangsters. That's how we knew the island's rules: 'Ja, you can stab me to death, but you'll go nowhere.'"

It is a strange story, isn't it? The gangsters' fears were surely unfounded. Apartheid was a callous system, but its prisons were not concentration camps, not even the jail in the middle of the sea. If an inmate were to vanish from the face of the earth, his family would ask questions. The prisons service would have to provide answers. On murder, at any rate, there were accounts.

But gangsters believed that if they stabbed, they would die. And their perceptions shaped prison life. The Number became a shallow echo of its usual self. There was no recruitment and there were no promotions. Nobody was hauled before inmate courts and no one was punished. 28s would get together and talk, but idly. For ndotas anxious to climb the ranks of their gangs, it was wasted time.

So Robben Island was something of an incidental control experiment. If the relationship between gangsters and warders is indeed framed by a structure of implicit rules, what happens when one side throws the rulebook away? What happens when warders announce—wordlessly, by the gesture of giving inmates knives—that Number ritual is punishable by death? The gangs evaporate. They become ghosts, living their frail afterlife on memories of what they did when they were flesh-and-blood beings. Indeed, the gangs' own metaphors described things this way. "In maximum, the Number is vleis en bloed. Robben Island is die stokkies" (limbo).

It is testimony to the massive gulf that lies between the camps of a totalitarian state, where people can indeed "disappear" with little fuss, and the brutal jails of a merely authoritarian state, which is still held to account. There are no gangs in Primo Levi's Auschwitz—a world where warders kill at will. However much the veterans of the Number may talk of the holocaust they witnessed under white rule, the apartheid prison was no concentration camp.

Chapter 3: The functions of violence II—Making men (and not women)

"It was my experience, as a soldier," Gore Vidal comments dryly of his sexual adventures in the U.S. Army during World War II, "that just about everyone, either actively or passively, was available under the right circumstances. Certainly, things were pretty open in the Pacific Islands, where on one, no doubt, mythical island an entire marine division paired off… Perfectly 'normal' young men," Vidal continues, "placed outside the usual round of family and work, will run riot with each other."33

For the past century, South African prisoners—placed outside the usual round of family and work—have indeed been running riot with one another. What has haunted them is not so much the sex itself, but an ambiguity that comes with the sex—the figure who is both a bandit and a passive sexual partner. It would be no exaggeration to say that sexual politics behind bars is devoted almost entirely to the (impossible) task of drawing a bold line between bandits and "women". The figure who is both a bandit and a woman is, it seems, unthinkable. He has to be erased in order to re-establish the order of things. Successive generations of inmates have poured their imaginative resources into making him invisible. They have erected a vast edifice of symbols, narrative and myth to do so.

At the centre of these symbols, narratives and myths is the figure of Magubane. He is the Number gangs' worst fear—the man who is also a woman, the bandit who parts his legs for other bandits. It is the ambiguity he represents that needs to be erased.

When I say "worst fear" I am not using hyperbole. Notice how the story of Magubane hijacks the Nongoloza myth and takes it places it was never meant to go. It began as a tale about the initiation of a group of anti-colonial rebels: they became bandits to defend themselves and their people from the whites' gold mines. Yet, no sooner has the band formed than it splits acrimoniously over whether bandits can sleep with one another. From now on, the heart of the story is no longer about fighting whites: it is consumed by the question of sex. Moreover, the quarrel can never be resolved: that much is built into the structure of the story. There was a man who knew right from wrong in this matter—Po—but he is dead. There was a document of founding principles, but half of it has been washed down the river. This is, perhaps, a begrudging acknowledgment that the problem is both fundamental to prison life and irresolvable, that it will never go away.

How do the gangs deal with Magubane, and thus with the question of sex among ndotas? Each gang reinterprets the Nongoloza myth privately and uses it to spread rumours about the other gangs. Each gang acknowledges that Magubane exists, but insists that he lives there, among that gang, not with us. So many of the fights and skirmishes between the gangs have boiled down to a single accusation: "You are Magubane."

The 27s and 26s insist that Magubane inhabits the silver line of the 28s. They say that the sole, or at very least the primary, function of silver-line officers is to serve as concubines for the gold line, for the soldiers. They themselves, the 26s and 27s claim, abstain entirely from sex when they are in prison.

28s—in particular silver-line 28s—deny this vociferously. They say that ndotas cannot have sex with one another, and that silver and gold-line officers are all ndotas. Rather, they argue, all senior officers, including senior silver-line officers, are permitted to have sex, but not with fellow 28s; they have sex with "sex sons". A sex son is a frans who lives among the 28s and sleeps alongside probationers in the silver line. But he is not an ndota. He will never be initiated. He will never learn about the Nongoloza myth, or about the rules and lore of the 28s. He is, they say, a sort of a prostitute. He is too weak to survive in jail on his own strength, so he turns to the 28s for protection, and in exchange he gives them sex.

But even this—sex between a 28 ndota and a sex son—is highly ritualised and meticulously clothed in meaning. The 28s consider themselves the children of Nongoloza, and as such they see themselves as Zulus—in war, in justice and in sex. In regard to sex, the relationship between an ndota and his sex son is understood as the relationship between a young, unmarried Zulu warrior and a Zulu maiden:

You know among the Zulus [a gold-line 28 told me], a young warrior and the girl he will marry are allowed to have sex before they marry, but only through the thighs, never penetration. Only after the wedding are they allowed to have real sex. It is the same with a sex son and an ndota. At night, the Silver Two takes the sex son to the ndota who has chosen him, say the Fyland. The Fyland has sex with him under the blankets, but it has to be face-to-face, never from behind, in order to show that they are having thigh sex, not penetration. At any point during the night, the judge is allowed to come to the Fyland's bed and lift the blanket. If he finds they are not face-to-face, the Fyland faces serious charges. His punishment is usually gang rape. And after that he is finished: because an ndota who has been raped is no longer an ndota.

Most 28s in the Western Cape are coloured men. They have made themselves into Zulus for the purpose of having sex with other men. Zuluness—a culture that many of them think of with superstitious distaste when they are on the outside—is embraced in prison because its strangeness helps them to set aside their own ambivalence towards what they are doing. In jail, they are Zulu men. They are having sex with Zulu maidens, with girls, through the thighs, the way they imagine Nongoloza and his forebears did. The line distinguishing men from women, the active from the passive partner, is drawn repeatedly, obsessively. To be the active partner on must be an ndota. In other words, one has to stab, one has to be beaten to a pulp without crying out, one has to sit in a dank cell for weeks and eat a saltless diet, one has to emerge from agter die berge strong. To be a woman, one must be nothing: a being who can never join the Number, who must walk barefoot and never leave the cell without permission, who must not conduct business in the public world of the prison. And as a reminder that things are not really so, that she is indeed a man, she must never be penetrated. There can be no real consummation.

Here, then, is another function of the violence of the initiation ritual. The first function, I argued in the last chapter, was to adjust to the infantilisation that attends on inmate experience: the ndota is a man and not a child. The second function is to make inmates men, rather than women. It is an attempt to rid the jails of that ambiguous figure who is a bandit and also a woman. So obsessive is the attempt to draw a line between the two that inmates end up engaging in the most vicious parody of the misogynist relationship between a man and woman on the outside. The active partner is steeped in violence; he is also the archetypal agent. The passive partner has been stripped of all agency—he is a helpless dependent.

The 26 and 27 understanding of the 28s—that the silver line serves the sexual needs of the gold line—is so ubiquitous, so widely believed, that at first I assumed it must be true.34 At first I thought that the silver-line officers who told me otherwise were merely concealing their shame. Yet the story is considerably more complex than that. The senior silver officers, at any rate, are undoubtedly not concubines; indeed, their role in the gang is quite an extraordinary one. There exists a sharp division of labour between the senior officers of the gold and silver lines. At one level this division is between intellectual work and the work of warfare. At another it is between the will to life and the will to kill. Here is a 28 silver-line officer's account of the relationship between the gold and the silver line.

We think of ourselves as Zulus, because our father, Nongoloza, was a Zulu. Think of a Zulu kraal, of how it is organised. The soldiers are not in the heart of the kraal. They are at the outposts, guarding the centre. That is the gold line. At the heart of the kraal are its brains, its memory, its very soul. That's us: the silver line. We are the Number, the keeper of its rules and its history. The soldiers are not the Number—they are its protector, its outer shield. When you recruit a soldier, you are not looking for a brain. You are looking for an aggressive maniac who acts before he thinks, someone who will put his life on the line instinctively, without asking questions.

This is, of course, a jaundiced position, one filled with silver-line chauvinism. Stripped of the chauvinism, however, it does in fact chime with the actual division of labour between gold and silver. The top two positions in the silver line, for instance—the mtshali and the nyangi—tell an interesting tale. The mtshali is entrusted with Rooiland's skin, and thus with the rules and the sacred history of the 28s. Indeed, he warehouses all the gang's accoutrements: its knives, its bayonets and its uniforms. He does so figuratively, rather than literally. Rooiland's skin is, of course, imaginary, as are the uniforms. But what this means materially is that nobody steps into his imaginary uniform, and thus his new rank, without the sanction of the silver line's leader; nor can soldiers arm themselves without his approval. And the nyangi, although not himself a soldier, is the one who inspects and issues weapons to the combatants in the gold line. Any soldier who uses a knife that has not passed through the hands of the nyangi has committed a crime. So the silver line is clearly a great deal more than a group of concubines—the gang can take neither judicial nor military action without its approval.

The top two positions in the gold line, by contrast—the judge and the general—are the man of the gallows and the man of war respectively. The judge is the member of the 12-man court that passes sentence. But he cannot pass a sentence without deferring to the judicial knowledge of the mtshali, to the rules on Rooiland's skin. The general is the leader of the combatants. It is said that he spends his days over a furnace shaping red-hot weapons with his anvil. But he cannot issue the weapons he has forged without the sanction of the doctor.

More interesting, perhaps, is the role of the silver line in the Twelve Points, the highest judicial structure of the gang, and the only one empowered to sentence an offending 28 to death. When the Twelve Points meets to decide a case, the judge, the most senior officer in the gold line, is the first to speak: he argues for a conviction. The mtshali, the most senior member of the silver line, rebuts him with an argument for acquittal. Then the forum debates the two positions. The judge ends the discussion by announcing the verdict. He then suggests a sentence. If the sentence is death, the mtshali immediately stands up and begins an argument in mitigation. The second most severe sentence in the 28s is a "band"—gang rape. The mtshali will generally argue that the death sentence be commuted to a "band".

Then everyone, save the general and the mtshali, votes. If five vote for death and five for a lesser sentence, the mtshali casts his deciding vote and the accused's life is saved. The judge accepts that his recommended sentence has been overruled. If those who vote for death are in the majority, the mtshali registers his protest by refusing to vote at all.

At this point—once the accused has been sentenced to death by a majority vote—the Goliat-one—the lowest-ranking silver-line member of the Twelve Points—steps in to save his life. He strips off all his clothes and runs naked round the edge of the Twelve Points; as he does so, he lets out a scream, in the most plaintive and haunting voice he can muster, pleading for mercy.

"The Goliat-one," a 28 told me, "is called The Man of the Light, the one who shines a light into the darkness of the 28s. To sentence somebody to death is a deed of blackness. The Number must ask itself: 'How did we get ourselves into this position? Have we become lost in the darkness?' Hearing the cry of the Goliat-one brings us back to the light. And if that does not work, if the hearts of the Twelve Points are not stirred by the Goliat-one, then the accused must be executed."

The accused is not present at the trial. He is not informed of his fate. After sentence is passed, word is whispered down the ranks of the gold line. Two or three rank-and-file soldiers are tasked with performing the execution. The most common method is to approach the accused in the middle of the night, while he is sleeping, and to suffocate him with a pillow.

"Which are the capital crimes?" I ask several 28s.

"Informing for the boere," they all reply, "and raping a fellow 28."

Interpreting the Twelve Points via its rituals, and in particular via the division of labour between its gold and silver members, it appears less a law court than a representation of an organism's unconscious life. The judge is compelled to kill, the mtshali to save. The accused's circumstances seem incidental—his case is merely an occasion for the gang to express its rival instincts. It is as if the entire structure of the gang is a device to render this conflict corporeal, to enact it on a physical stage.

The silver line appears to be the gang's life force, its instinct for preservation, and it is eternally pitted against the death drive represented by the general in the gold line. The mtshali is the life force's intellect, its capacity to reason. The Goliat-one is its raw, naked heart: his anguished scream expresses the desperate urgency of the desire to preserve, the unspeakable horror of death. The relationship between gold and silver takes on the form of an existential rivalry, a battle over life and death.

So much for the stories that the 26s and 27s whisper about the 28s, and the counter-stories the 28s tell in their defence. What do the 28s say about sex in the 26s and 27s?

As with so many others matters, they refer back to the Nongoloza myth. Only, they give it their own private ending.

The story begins with a question. What happened to Magubane? Recall that after Kilikijan found Nongoloza making love to Magubane under a cowhide, the bandits split into two groups. Magubane changed allegiances. He went with Nongoloza's group, making it eight, leaving Kilikijan with seven.

Both sides agree what happened next. Nongoloza's men continued their work of robbery and plunder for some time. Magubane stayed with them a while, just long enough to win their trust. Then he used the trust he had earned to con them. He volunteered the responsibility of safekeeping a large proportion of their booty, and then absconded with it and disappeared.

That much the 26s, 27s and 28s agree upon. But the 28s add a final twist to the story. "What happened to Magubane after he absconded?" they ask. "Where did he go? Why does he never appear in the story again?"

They smile cynically. "Remember Grey," they continue, "the founder of the 26s, the leader of the six men Kilikijan found in Point Prison. What did Grey do? He made money by conning and tricking. Isn't that exactly what Magubane did to the 28s?"

Then they shrug and say no more. The silence finishes the story: Grey is Magubane. The 26s break their own taboo. They have sex with men.

Anyone who has served time in a 26-dominated cell knows that this is true. It is barely concealed. What is interesting, though, is the way the 26s explain it away.

Vincent Shabangu, the activist convicted of public violence whom I quoted in Chapter 1, spent several months in a 26-dominated cell in Pollsmoor's maximum-security jail during the course of 1986. He told me the following story.

"I was with the 26s on C section for a while, and then I was put in isolation because I was studying. In isolation, I met two other guys who had been convicted of public violence. One was called Gerald.35 I was moved to isolation on a Friday afternoon; he was moved on the same Friday night. It seemed that something terrible had happened to him on C section. He cried all weekend. He was really frightened. I coached him. I told him that if he keeps to his faith, he'd be okay.

"Unfortunately, I soon discovered, he had been drawn in. He had become a wyfie to feel safe."

"A wyfie in C section?" I asked. "A 26's wyfie?

"Yip. Even though they said they did not have wyfies. Even though they claimed that only the 28s have sex. They all had wyfies. Let me tell you how I found out about Gerald. The 26 general in C section decided that there were certain weekends when Gerald must go to his cell to watch movies. One Sunday, a warder who works in C section came round to fetch Gerald, and I said I also wanted to go and watch a movie. Why shouldn't I be allowed to go too? When I went, I saw they were making Gerald a special bed next to the general's bed. It was so strange to see. They were making preparations for the general to have sex with him. And it wasn't just Gerald. Two other activists also became wyfies to 26s. If you are not strong, if you are not well prepared, they corrupt your mind, get into your mind."

"But the 26s and 27s hate the 28s," I say. "And they hate them because the 28s have sex."

"Yes," he says with excitement. "They claim they are not having sex with the guys. They don't call them wyfies. I do. They claim they are protecting the young guys from the 28s. They say: 'Look at that ndota from the 28s. His eyes are all over you. You need me to protect you from him.' They even said to me, right at the beginning: 'Don't mix with 28s. If they get a chance, they will rape you.' But that is exactly what the 26s were doing. They have the young ones with them. They are having sex with them.

"And let me tell you something interesting. For a one-week period, just before I was put in isolation to study, I was taken to E section—the 28 section of the prison. What was happening in the 28 section was different. There was a different atmosphere. There was sex happening there too, but you didn't feel frightened at night. You would go to sleep in your little corner and know you were safe. Not in C section, not with the 26s. There, things were unpredictable. They could decide to do things in the middle of the night.

"I told the 28s. I said, 'On C section, the 26s say the youngsters have to be protected against you.' They laughed. They said: 'The 26s are hypocrites: the youngsters need to be protected against them, not us.'"

Everybody accuses everybody else of being Mugabane. "Yes, Magubane is here, but not with us. He is with them." Yet Magubane is everywhere—in the 26s, 27s and 28s. In the following section I show that even in the 28s, who have managed the dilemma of homoerotic sex far better than the other two gangs, the task of constantly retracing the boundary between men and women, the initiated and the passive, is an uphill battle.

Number-gang violence, I have argued thus far, performs at least three critical functions. First, it makes inmates into men, rather than boys. Second, it is used to patrol the boundaries of gang space against warders, the boundaries within which men can be men. And, thirdly, it divides inmates into men and women, thus masking the intolerable ambivalence of the bandit who is also a passive sexual partner.

In the previous chapter, I pointed out that the violence to which ndotas subject themselves is as critical in the making of men as the violence they commit. The beatings, the solitary confinement—these are pivotal in the forging of Number-gang virtues. The violence to which they subject themselves is predictable and ritual, and they court it openly, but it is severe nonetheless: broken bones, permanent scars. In the previous chapter, I raised, but did not answer, a question: is the price of making men too high? Are the funds of ndotas' stoical indifference to physical pain limitless? In this section, I explore these questions.

Something of great import happened in the Number gangs of the Western Cape in the late 1980s: they "closed their blood lines". In other words, they decided they would no longer stab.

The story is a murky one. There is no written record of what happened, as far as I know, only hundreds of memories. Each ndota I have interviewed tells a rival tale: where and when the decision was taken, how it spread so quickly and, above all, why it happened. It is probably too late for a viable and comprehensive account to be written now. A decade and a half has passed and the memories of veteran ndotas, it seems, grow less reliable with time.

In any event, it was nothing less than a cataclysmic decision. To stab, and to be subjected to violence and deprivation in the wake of stabbing, was the centrepiece in the initiation process. It animated the world behind bars, providing it with its fundamental meaning.

Why did it happen? I don't know. Some of it may have had to do with the changing relationship between prison gangs and street gangs, which I discuss in the following chapter. Some of it may have had to do with changing conditions of imprisonment. If so, I have struggled to put my finger on what, precisely, changed: the early 1990s would be characterised by widespread uncertainty and far-reaching reform in the ranks of the prisons service, but not the late 1980s.36

There is one common denominator in ndotas' accounts of the closing of the blood lines: the decision to stop taking blood was made by the soldiers themselves, by those whose business it was to stab, and then to be punished by the authorities for stabbing. Who were they? Two groups, primarily. First, the 27s, whose job was both to defend the 26s and to "right wrongs" between the 26s and the 28s by taking blood. Second, the soldier line of the 28s, who used violence both to defend the gang from attack and to draw the authorities' attention to poor prison conditions.37 In other words, it appears that the decision was taken by those in the firing line, those who had to endure, in the course of their business, submission to violence.

Indeed, the position of the 27s has always seemed puzzling, almost unfathomable. I said earlier that the virtues instilled by initiation are restraint, stoicism, selfless solidarity. The very fabric of the Number gangs is permeated by personal denial. But in the 26s and 28s, this denial is offset by carefully circumscribed forms of excess. The 26s are accumulators of wealth—one is rewarded by material comforts. And a soldier in the 28s is rewarded by creature comforts. Indeed, the gang says as much. A soldier's catamite, the 28s say, is his reward for going to battle.

But the 27 is truly a Spartan figure. There is no reward for him other than the power and mystique of his position. He is not allowed sex and he does not engage in commerce. Aside from learning the law of all three gangs, his sole function is to stab. He thus commits himself to a life in prison as his time escalates for the serial crimes he commits on the inside. And, through the course of his career, his body is subjected to serial battering. Of all the positions in the gangs, he is the impossible figure.

And indeed, it appears from the records of commissions of enquiry, court cases and warders' testimony, that 27s have often, from the earliest times, been on the brink of extinction.38 It also appears, from inmate testimony, that there have always been pretenders who have attempted by guile and finesse to occupy the place of the 27s without earning it.

In any event, it is interesting that all 30 or so of the veterans I interviewed agreed that it was the soldiers themselves who instigated the closing of the blood lines. According to a 26 with a flair for telling stories:

It was the 27s who did it. You had to have been there and seen them to understand why. They had been too badly fucked up. Some were on crutches; others were so damaged on the inside that they couldn't eat properly. Others were just miserable inside their heads. They were a sorry bunch. They had to do it to keep their own dignity.

A silver-line 28 told me that:

The soldiers in the 28s just laid down their arms. Their work was too tough for them. They couldn't handle the carry-ons and the one-ones. Too many of them were crying. Too many of them were nursing their wounds. They destroyed the Number because they lost their courage.

Whatever the reasons, the closing of the blood lines caused havoc. Some time in 1987, either just after or at the time of the decision to stop ritual stabbing, a province-wide conflict erupted between the silver and gold lines of the 28s. Once again, the circumstances are murky. It appears that the conflict began in Victor Verster Prison and soon spread to prisons throughout the Western Cape. The veteran 28s I interviewed give rival versions of why the war started. Some say that the silver line rebelled against the decision to end ritual stabbing. "The soldiers weren't soldiers anymore," a silver-line 28 told me. "They were not doing their job. They were not defending the camp. So they were useless to the camp. We declared war against them because we wanted to overthrow them."

According to others, it was the issue of food that precipitated the conflict. There was an old rule in the 28s. The deprivations the gold line had to suffer went beyond subjection to ritual violence. They were also "not allowed to get fat". They were not allowed to eat meat—they could only have the bones. They weren't allowed to eat eggs—they had to give their eggs to the silver line. No sugar in their porridge; they had to have bitter porridge. No sugar in their coffee.

"Now that we had stopped taking blood," a gold-line 28 who was serving at Victor Verster at the time told me, "we demanded that we should eat: the meat on the bones, eggs, sugar. It was time for the silvers to start sharing the food. They said: 'No. The meat and eggs are ours.' So we thought: 'Fine. We will do to them what they have been doing to us. Soldiers started guarding the food trays. We wouldn't let the silvers near the food. We were starving them. For three days, they did not eat. Then they declared war on us."

The third explanation for the conflict is perhaps the most interesting. According to some, soon after the decision to end ritual stabbing, low-ranking soldiers began raping low-ranking silvers. Senior silver officers complained to senior gold-line officers, presuming that the junior soldiers would be punished. Instead, the gold line stood by and did nothing. The silver line then declared war.

(It is possible—although this is highly speculative—that soldiers had always had sex with junior silvers, despite the prohibition; that it was only after the decision to end ritual stabbing that the silver line began prohibiting sex with its junior members. In other words, it is possible that it is only once men were no longer men, so to speak, that senior silver-line officers began objecting to soldiers having sex with junior silvers.)

Whatever the causes of the conflict, there is consensus among those I interviewed about the outcome. The silver line trounced the gold; the soldiers were beaten by the non-combatants. Within months, the conflict was over. "The hospital beds were filled with 28 soldiers," a gold-line 28 told me. "Some of us were walking round with bandages on our heads. Some had broken limbs. I myself had a broken collarbone. We said to the silvers, 'Okay, that's enough. What do you want?"

What the silvers wanted, according to the legend that survives today, is even more extraordinary.

We closed down the gold line [a silver-line officer told me]. And we vowed that it would never open again. We threw the soldiers out of their offices. We chose people from our own ranks and filled the gold line with them "just for the minute".39 We said that nobody would ever have to take blood to join our camp again.
And then we did something really stupid, something so corrupt and evil that we are still living with the consequences today. We changed the initiation rules. We said that from now on, everyone who joins the 28s must have a "babba". In other words, the ndota who trains you has sex with you. To join the 28s, you have to be a sex son first. And so it happened. The sex sons became 28s.

I say "according to the legend that survives today" because the real story—untainted by the distortions of memory and probably lost forever now—is surely a good deal more complex than this. It is unlikely that initiation into the 28s through sex occurred as an event, as a single rupture. It is much more likely that the rumours the 26s and 27s tell about the 28s have always contained an element of truth, that there has always been a degree of ambiguity in regard to how one joins the 28s.40

Nonetheless, even as a product of memory, the story is a telling one. For their entire inmate careers, 28s had had to withstand jibing and rumours from the other gangs—a kind of whispering campaign. They had been told that, no matter what they said in rebuttal, they were the ones who had sex with one another. Now, in the aftermath of a bitter conflict, the 28s make real the very rumours against which they have railed. They institute as law the idea that one must become a passive sexual partner to join the gang.

There are many things to be said about this strange story. One of those things, surely, is this: the moment the blood line closed—in other words, the moment the manufacturing of men ceased—so the line dividing men from women was erased. The entire edifice of 28 myth was a careful, painstaking attempt to separate ndotas from "women". Now, it all fell apart. With the new form of initiation, women and ndotas began to become indistinguishable.

In the 1990s, the blood lines never really reopened, but they never really closed either. They never reopened in the sense that most inmates today who are recognised as soldiers did not stab to acquire their positions. According to the veterans—those who cut their teeth in the 1970s and 1980s—1990s soldiers are mere boys: they are soldiers "just for the minute". Nonetheless, whenever there is a crisis of authority in the gangs, whenever two men or groups tussle for power, the knives are taken out again. All of a sudden, the question of who is "meat and blood" and who "just for the minute" becomes salient. In others words, the lost tradition is shored up whenever there is a legitimacy crisis.

But to understand the word of the 1990s and 2000s in its proper context, it is necessary to turn to the relationship between street gangs and prison gangs. This relationship underwent something of a revolution in the early 1990s, changing the face of both the prison cells and the streets forever.

Chapter 4: Prison on the streets, the streets in prison

When an unknown inmate is transferred to a new prison, he is asked a question the moment he enters his cell: "Who are you?" The question begins a test of authenticity. If he is a 28, he must sabela a reply that he is a son of Nongoloza, that he works by night. If the questioner is a 27, he will respond by declaring that he is Kilikijan, that he works by day. The two then run through the motions of a formulaic and well-rehearsed exchange, together describing the early history of Nongoloza and Kilikijan. The new arrival must use the right words and the right metaphors; everything hinges on his mastery of prison language. When Kilikijan is satisfied that the new man is a genuine child of Nongoloza, he delivers him to the glas of the 28s, who asks him the same question: "Who are you?" Now that he is in his own camp, the new arrival must describe each rank in the 28s. "The glas blows his bugle, the nyangi throws his pipes, the general sits with his back to the four points of one-times…" He must describe every rank except his own; he identifies himself by his omission.

According to several of the veteran inmates I interviewed, something strange began happening at the end of the 1980s. A new inmate would arrive in a cell. He would be asked: "Who are you?" The newcomer would answer confidently that he was, for instance, a soldier in the gazi line of the 28s. But the manner in which he answered, his clumsy use of prison language, his haphazard grasp of the Nongoloza myth and the terminology, demonstrated that he was not, in fact, a 28, that he knew next to nothing about prison gangs.

The inmate would be quizzed further.

"Where did you become a 28?"

"In Hawston"—a small fishing village about 100km east of Cape Town.

"But there is no prison in Hawston," the ndota would say.

"I was not recruited in prison," the newcomer would reply. "I was recruited by the Rooidakke"—a prominent street gang in the late 1980s and early 1990s. "I was ordered to do a hit, to shoot somebody. They said that to join the gold line of the 28s, you have to take blood. Once I'd done the hit, I would be a 28."

The inmate would reel in astonishment. He had never heard anything so crazy in his life.41

Why did the rubric of prison gangs begin spilling onto the streets? It is difficult to answer this question, except in speculative terms. Ever since the 1970s, there had been a loose and informal relationship between prison and street gangs. Two of the major street gangs of the late 1970s and early 1980s were the Scorpions and the Born Free Kids. When they went to jail, Scorpion members generally joined the 28s. (It was said that a Scorpion must go to the 28s to fetch the poison for his tail.) Born Free Kids generally joined the 26s. (It was said that both 26s and Born Free Kids were "chicos"—smooth dressers.) But it was a loose relationship. It was understood that the walls separating the prison from the street were sacred, that there was no such thing as a practising 26 on the outside: the very idea was absurd.42

All this began to change in the late 1980s. The imagery and ritual of the prison arrived on the street—bastardised, in scraps and pieces—and it spread like wildfire. By the late 1990s, two of the major street gangs of the time—the Americans and the Firm—had adopted Number-gang ritual wholesale. (The Americans adopted 26 ritual, while the Firm adopted 28 ritual.) Indeed, in the early 2000s, the Firm began calling itself the 28s. Its leaders had designated themselves generals, and had appointed captains, sergeants and judges.

The street gangs took the world of the prison—its metaphors, its nomenclature, its logic—and imprinted it on the ghettos. Ever since the early 1970s, the street gangs of the Cape Flats have been extorting protection money from neighbourhood shops, demanding a cut of liquor distributors' profits, taking transit fees from the taxis that drive through their turf, maiming those who dare to sell anything without their permission. But now, beginning in the 1990s, street gangs began using prison as a metaphor to understand their relationship with those upon whom they preyed. The street gangsters are the ndotas; the taverners, liquor distributors and taxi drivers from whom they extort are the franse. Like the franse behind bars, they too must rent the air they breathe.

It is a strange thought, isn't it? For much of the 20th century, inmates imagined the jails that housed them as the open plains of the 19th-century highveld and the forests of early Natal. Joining the 28s, for instance, was described as a Homeric journey through a wooded wilderness. Now, in the late 1990s, the youths of Cape Town's ghettos began to imagine their neighbourhoods as prisons, each piece of turf a massive jail cell of the initiated, every taxi owner a frans to be milked.

The street gangs had finally "stolen" prison: they had turned the institutions that punish them into founts of inspiration.

It is not just the supergangs. The inspiration of prison has permeated the most parochial street corners. In 2002, in Mitchells Plain, I met a 21-year-old man who had joined a gang unheard of outside his neighbourhood—the Jolly Killers. I asked him to do a piece of research work for me and he agreed on condition that I pay a third of his fee in advance. A week later, I phoned him to ask how the work was going.

"I'm not going to do it," he replied. "I'm a 26. My work is to con you out of your money."

"You're a fool," I said. "It wasn't much work and if you'd done it you would have earned a whole lot more."

He laughed patronisingly. "You don't understand. I'm a 26. That's my ethos."

"Who made you a 26?" I asked.

"The leader of the Jolly Killers went to jail and became a big 26," he replied. "When he got out, he recruited us all."

Why, after so many years, did it happen when it did—in the late 1980s and early 1990s? I can only give a partial and speculative answer. In the first place, the character of Cape Town's underground economy had begun to change substantially by the late 1980s. The major gangs of the previous decade—the Born Free Kids, the Mongrels, the Scorpions—all had a regional presence in the Western Cape, but they were, in reality, little more than regional affiliations of local groups. Gang leaders made money, but not that much money. They controlled turf, but not that much turf. The typical gang leader would control a few dozen blocks of a ghetto. He would run its mandrax and marijuana trade, much of its liquor trade, extort protection money from its shopkeepers, control its commercial sex industry, buy and sell stolen electronic equipment. The underground market was static. It was confined largely to mandrax, marijuana, liquor and stolen goods industries, and to small-scale extortion.43

The gang leaders who came of age in the late 1980s did so in a very different world. The rapid insertion of South Africa into global markets brought new drugs into the country (crack, heroin, club drugs), new merchants and, above all, new markets. An impressively innovative network of West Africans took over inner-city prostitution and used it to create a brand-new crack market, one that, significantly, cut across class and racial lines. Also, the emerging rave scene in South Africa brought with it a host of new club drugs, bringing scores of middle-class youths into the drug market.44

The traditional street gangs of the Western Cape found not only that drug consumption was growing rapidly, but also that a sizeable proportion of the emerging drug market was located outside their traditional turf. With the prospect of their own clients deserting them for new drugs and new players, they knew they had to expand quickly or risk being swallowed.

The major gangs scrambled to consolidate their traditional constituencies and to lay down new turf throughout the Western Cape: from the inner-city suburbs of Cape Town, to the coastal villages both east and west of the Cape Peninsula, to the rural towns of the Western Cape hinterland.

It was a time of great risk and uncertainty for Cape Flats gang leaders, but a time of great reward for those who succeeded. The late 1980s saw the emergence, for the first time in the history of the Western Cape underworld, of stinking-rich Cape Flats men, millionaires many times over: people like Colin Stansfield, who cut his teeth in the Scorpions and scaled the ranks of the 28s; Ernie "Lastig", who led the Rooidakke and is today reputed to control the lucrative poaching of abalone off the southern Cape coast; Jackie "Lonti", a 26, leader of the Americans before his assassination in 2000, and the man reputed to have brought crack cocaine to the Cape Flats.

Men like these were the first in the history of Western Cape crime who needed to build and maintain province-wide allegiances. They required the allegiance, not only of foot soldiers, merchants and street dealers (it takes a veritable army to control a province-wide trade) but, above all, of consumers. A large portion of Western Cape drug consumers had always bought their drugs according to their gang allegiance. Now, the allegiance would have to be to a large, abstract, province-wide entity. In other words, the imagination of consumers became a vital resource on which to work.

What better set of tools with which to reshape the imagination of marginalised young men than the iconography of the prison? For decades, rumours, tales, slivers of narrative about the Number gangs had been trickling onto the streets. Prison generals walked out of jail demigods. The words "28", "27" and "26" had long been sprinkled with magic. The real question is why Number lore had not taken hold on the streets earlier. Indeed, the way some interviewees describe it, the introduction of Number lore on the streets seemed organic, natural: it captured the imagination immediately.

In mid-2003, for instance, I interviewed a gangster who had been one of Jackie Lonti's most trusted lieutenants in the late 1980s, precisely the time when the Number found the street. I asked him if he recalled the very beginning, the first time the Americans began, incipiently, to understand themselves as 26s. He replied:

I'm not good with dates, but sometime back then, in the 1980s, Jackie spent a while in jail. When he came out, he brought the Number out onto the street. To deal now, you had to know prison gangs.
I remember once, a few days after Jackie was released, there was someone dealing in buttons in Athlone who Jackie didn't trust. He said it was time to get rid of him. But he didn't say it like that. He said: 'No, that guy is a frans; he can't work with money.' He called this guy to his house and said to him: 'Hei, wie is jy?' ('Hey, who are you?')—just like in prison.
It was strange when I heard that. On the one hand, it was new—nobody had ever said that on the street before. On the other hand, it felt like it was old, like we had always thought that way; only now, Jackie had put the right words to it.
It was received on the streets as if it were organic, as if had always belonged there. But there can be little doubt that it was introduced as a conscious strategy. Faux Number gangs—prison gangs on the streets—were the thread with which the likes of Lonti stitched together regional empires.

So, in the late 1980s and early 1990s, young men arriving in the prisons began to announce that they had been made ndotas on the outside by men like Lonti and Stansfield. These men were famous and rich, and the veteran ndotas behind bars knew it. But they also knew that such men did not hold senior rank in the prison gangs proper, that they did not have the authority to initiate people on the inside, never mind on the streets. So, many among the first generation of the new "street ndotas" were ostracised and beaten when they came to jail.

But that did not last long. As the likes of Lonti became multi-millionaires, their cachet grew. They became heroes, indeed gods, to scores of young men of the Western Cape underclass. Some of them crafted their own heroism by deftly inserting a subtle political subtext into their discourse. They started off as dirt-poor ghetto kids under a racial dictatorship, they said, and they had become the richest and most powerful men in the province in spite of that. They had defied apartheid, not just by breaking its laws, but by becoming omnipotent. Indeed, they were far more competent than the Number gangs ever were at blurring the distinction between crime and politics, at spinning banditry as a life of political virtue.

At this, Colin Stansfield was the best of all. When the African National Congress won South Africa's first democratic election in April 1994, he appropriated the victory as his own. He hired the local football fields in the ghetto of his childhood, Valhalla Park, erected a giant marquee, bought thousands of litres of beer and truckloads of meat, and invited every resident of the ghetto to attend. He threw the biggest party in Valhalla Park's history.

Against all this, the old tradition of orthodox prison gangsterism stood little chance. Many of its proponents were crusty old-timers, rank with the smell of many years in prison. Who were they in the face of Stansfield, Lonti and Lastig—the demigods of the new era? It was not long before the street gangs had begun not just to adopt but to appropriate the world of the Number as well. It was not long before those initiated on the outside were indeed ndotas, because what "ndota" meant now was totally new.45

Several interviewees pointed to a specific event in 1996 in Pollsmoor Prison as a turning point. Jackie Lonti was interned at Pollsmoor. He was, at the time, among the two or three most powerful men in Cape Town's underworld; the gang he led, the Americans, was probably the largest and the most prosperous. He was a 26, but not a general. Many of the 26s around him were also Americans. So, they were his followers on the outside, many his equals or seniors on the inside. Among the 28s in Pollsmoor at the time were leaders of The Firm, the Americans' major rival in the region's illicit economy, and thus Lonti's enemies. According to one interviewee, a 26 and an American:

Jackie sent out an order to all 26s in Pollsmoor: he told us to raise the Star-Spangled Banner in our cells, and to declare war against the 28s. You must understand that what he did was so wrong according to our tradition that he should have been sentenced to death for it. First, he told us as 26s to raise the flag of the Americans, a vuil papier, a flag of the fourth camp. That alone is punishable by death. Second, he declared war against the 28s when the issue that sparked the war was an issue from the outside, not from the four corners. That alone is punishable by death—you do not bring the outside into prison. Finally, he was not a general. He did not call a meeting of the Twelve Points. Only they can declare war against the 28s, and only in consultation with the 27s. That, especially, is punishable by death.
But he was Jackie, and Jackie meant more than anything else. So we raised our American flags in our cells and we went to war with the 28s. Something like that had never happened before. After that, the Number was never the same again.

When I arrived at Pollsmoor Admission Centre towards the end of 2002, the prison was in flux. There were very few high-ranking street-gang leaders in the prison at the time, and thus little authority. A host of pretenders were jostling for power in all three gangs. It was a good moment to come to Pollsmoor, for the politics of a power vacuum is always instructive; in particular, the fraught, pained relationship between the old and the new, between the dying, yet still potent, traditions of the 1980s and the new ethos of the 2000s, was stripped bare.

Among the hundreds of 28s in Pollsmoor Admission Centre at the time, seven had been recruited in the 1970s and 1980s and, thus, in "the old way". They had all, they claimed, taken blood, been beaten by apartheid warders, spent time in isolation on a saltless diet. In the weeks after their respective initiations back in the 1980s and 1970s, they had all, they claimed, been taught the Number the proper way: taught to sabela, taught the history of Nongoloza, the rank structure, the rules. They, they said, were the only real ndotas in the prison, flesh and blood. Everyone else, they said, had been recruited in the 1990s and 2000s, after the blood lines had closed and after the street gangs had infected the prison. The others were all tronk laaities—mere boys, not ndotas. The term tronk laaitie is particularly stinging: it not only signifies a boy, rather than a man, but also subtly implies that its bearer has entered the gang by becoming a passive sexual partner. Tronk laaities, the seven veterans claimed, had not taken blood, or had cried when they had been beaten. Their knowledge of the Nongoloza myth was scrappy and piecemeal—they had picked it up off the streets; their knowledge of the gang's laws was so incomplete it was dangerous.

The veterans' talk really was as it sounds: a parody of a bunch of old men complaining about the youth of today. The result is that they talked more freely with me than anyone else did. They had reason to. They felt that the proud tradition that had once been theirs was being spat on. They wanted the past—or their idealised memories of it at any rate—to be recorded: they liked me for my notebook.

None of these veterans were particularly high-ranking. All were officers, but middle-level officers. All had acquired their current rank in the mid or late 1980s. They had all, they claimed, refused to take a higher rank after the blood lines closed because, ever since then, a higher rank had become meaningless—it was not flesh and blood.46

Six of the seven played only a marginal role in the day-to-day politics of the 28s. Most were elders, or "in die wolke" ("in the clouds'), as they say in the Number gangs—acknowledged, but inactive. Nonetheless, there were particular times when the veterans were to play a pivotal role. When and why tell us much about the politics of the Number gangs today.

In the remainder of this chapter, I track three incidents that happened while I was at Pollsmoor. The first was a conspiracy to stab, the second was the arrival in Pollsmoor of Colin Stansfield, one of the most famous gangster capitalists of the Western Cape, and the third a successful stabbing. All three incidents were characterised by a crisis of authority. In each, the project to shore up authority took the form of a failed attempt to invoke the rules of "the old days", the times when "ndotas were flesh and blood".

Incident 1: the failed recruitment of Rashid

In late 2003, a prominent member of the Americans—I will call him Rashid—was arrested for possession of crack cocaine and interned in the awaiting-trial section of Pollsmoor Admission Centre. At the time, Rashid controlled the Americans' operation in the Cape Town inner-city suburb of Sea Point. He was, in other words, very rich and very powerful. Being an American, he naturally gravitated towards the 26s. He was, however, an anomaly: despite his position in the Americans he had never been to jail and had thus never been initiated as an ndota.

When Rashid arrived at Pollsmoor, the 26s in the prison fell into a dispute. Faction A argued that he should immediately be given a middle-ranking position in the gang. Some said that his position in the Americans gave him automatic membership of the 26s, that to deny him membership would be to dishonour him. Others were more transparently expedient: Rashid had access to a large drug market, and the 26s knew how to get his drugs into prison, where they could be sold and thus enrich the gang.47 Giving him immediate membership would thus bring immediate reward.

Faction B in the 26s was outraged. Access to manhood cannot be bought, they said. A man in prison earns his manhood with violence. Rashid's position in Sea Point entitled him to nothing on the inside.

For days, tension simmered in the 26s. Faction B threatened that if Rashid was made an ndota, he would be stabbed. Faction A retorted that if Rashid was stabbed, the leaders of Faction B would all be killed. Finally, the dispute was taken for mediation to die manne in die wolke (the men in the clouds), the old-timers who had been recruited back in the days when the line dividing men from boys was clear.

Their decision was nothing less than extraordinary, an exercise in pragmatism so transparent it was almost amusing. Of course Rashid would have to stab to become an ndota, they said. How could it be otherwise? Moreover, since there was so much controversy surrounding his recruitment, he would have to stab in dramatic fashion. The Minister of Correctional Services was due to visit Pollsmoor later that month. It was to be a grand affair. A marquee was to be erected in the central courtyard. A host of dignitaries was coming to the prison. Rashid would have to stab another inmate there and then, in the middle of the minister's speech.

Now, it was quite clear to everyone in the 26s, including the men in the clouds, that Rashid was not going to stab anybody. He was, above all, a businessman, and a very successful one at that. Unlike most 26s, he had a high-calibre legal team on his case; he was expecting to be released on bail within days or weeks. If he stabbed another inmate in broad daylight, he would not only lose his bail hearing: he would be charged and convicted of assault with intent to do grievous bodily harm, and would spend years behind bars. He was not going to do it. It would be bad for business, to say the least.

So, the men in the clouds twisted the tail of their verdict. In deference to Rashid's prominent position in the Americans, they concluded, he would not have to commit the stabbing himself. It would be done on his behalf by two junior 26s, and they would be promoted to officer rank for their troubles.

Naturally, this did not satisfy Faction B's complaint. They nodded their consent—one does not openly challenge the men in the clouds—but hours after the meeting ended, its decision was leaked to a senior warder. He, in turn, kept his newfound knowledge close to his chest until the day before the minister's visit. That evening, he transferred the two appointed stabbers, as well as the entire leadership of the 26s, to another prison. On the day the minister came, every other prominent 26 was kept locked in his cell. Two days later, when the minister was safely back in Pretoria, those who had been transferred were brought back to Pollsmoor.

What had happened, of course, is that the 26s had attempted a clumsy marriage between two conflicting conceptions of value, status and authority. The first conception is animated by the world outside: it is a world of high-stakes illicit commerce, a world of multi-million rand businesses. Its protagonists are money men who have earned their authority through their wealth and, in turn, through their capacity to build and control private street armies. These are men who have borrowed the iconic status of prison and prison gangsters to organise their soldiers and their consumers on the outside. But their world is not prison—it is the illicit economy.

The second conception is animated by a century-long tradition behind bars. It is about the making of men in total institutions, a world sealed off from the outside, a world in which the purposes of ritual are to arrest the slide of inmates into childhood and womanhood.

For the duration of my research in Pollsmoor at any rate, neither conception of authority eclipsed the other. They co-habited uneasily, and the result is that both were disarmed. Inmate life, in other words, was characterised by a profound legitimacy crisis. The gangster capitalist of the world outside was unable to win hegemony in the prison, for the simple reason that the prison is, after all, the prison, and not the outside world. The making of ndotas through violence serves deep, visceral needs—needs peculiar to inmates—and they cannot be wished away.

So the gangster capitalists—in this case, Rashid—had to go through the motions of submitting to the ostensible authority of the men in the clouds, men Rashid has the right to treat like dirt on the outside. And, yet, the old men's authority was only ostensible because they knew they could not order Rashid to stab: he was, after all, an important man on the outside, and thus an important man in prison.

The result is that both rivals were defeated. Rashid was unable to play his role as the inmate man—an initiated ndota—and thus spent his brief time in prison in limbo: a big shot, but at the same time a little boy. Yet his defeat was hardly a victory for the traditionalists. The paralysis his stay in Pollsmoor caused only brought home that the rituals that so painstakingly, so precariously, separate men from women and boys, were in ruins. One way of expressing the problem is this: ever since Western Cape gangsters started becoming multi-millionaires, Nongoloza has been split in two. Out in the world, he is the bandit turned robber baron. He has used his outlawry to build an empire, and he has become the hero of the region's criminal classes. Inside, he is still the figure who builds the dam wall between men and children, and between men and women. Western Cape inmates cannot have both Nongolozas, but they cannot give either of them up. They try to piece together a composite Nongoloza—bits of one, scraps of the other—but the composite figure cannot stand on his own two feet.

Incident 2: a robber baron comes to Pollsmoor

In February 2002, Colin Stansfield, one of the first multi-millionaire gangster capitalists of Cape Town and a consummate populist, was sentenced to six years in prison for tax evasion. After sentence was passed, he was taken directly from the courtroom to Pollsmoor in order for his case to be processed in the Department of Correctional Services bureaucracy. A few hours after he arrived, he was whisked off to Helderstroom Prison, about 100km away, where he was to serve his sentence.

Stansfield was the founder of The Firm, which, by 2002, had begun calling itself "the 28s".

I was in Pollsmoor on the day Stansfield briefly appeared there. About an hour after his arrival, a rumour quickly spread through the prison.

"Colin arrived at the processing centre wearing a huge trench coat," I was told. It was late summer. The temperature was well into the 30s. "All the inmates in the processing centre surrounded him, and then he opened his trench coat: it was lined with R200 notes. Before the warders could get to him, he dished out the money to every 28 in the room. Every 28 got at least two R200 notes."

Later in the day, I found two warders who had been in the processing centre at the time. The part about the trench coat, it turned out, was fabricated. But the part about the money was true.

As the story spread through the prison, the 28s began speaking to one another about it. They were, I discovered, furious with Stansfield.

"Colin was only here for a few hours," I was told, "and he still had to buy his safety. He was frightened. Each of those R200 notes saved his skin. Because he knew that here he would encounter real 28s, real ndotas, and they would not be pleased with what he has done to our camp. On the outside he calls himself a general; in prison, he was never vleis en bloed. On the outside, he allows young boys to call themselves 28s, and when they get here, they think they're ndotas. Here, there are tough questions he has to answer. So he bought his way out. He is a weak man."

I doubt whether Stansfield's life was in danger for the few hours he was at Pollsmoor; I doubt whether anyone would have dared to stab him. The story I was told was spun from bitterness and envy, not from a sober reading of Stansfield's status in prison.

It is, nonetheless, an interesting story. It is a graphic depiction of the uneasy co-habitation of the gangster capitalist's ethos and the prison's ethos. The essence of the story is that men like Stansfield have stolen something with their R200 notes. They have stolen an inmate tradition. Its function was to create a world of meaning within the total institution, to sift through the complexity of inmate experience and to make something intelligible from it.

Incident 3: the stabbing of Warder Davids

In March 2003, a senior warder in B section, which houses sentenced prisoners, was ambushed in the corridor by three inmates and stabbed 12 times. I have called him Warder Davids. Most striking about the attack is that the aim was to kill. No short blade, no single wound below the shoulder, but a vicious, potentially lethal assault. As it happens, Warder Davids was not critically injured. But he did take several weeks off work, and the rest of the staff who remained behind were deeply traumatised.

In the days following the assault, the tension in B section was almost unbearable. The inmates were confined to their cells 24 hours a day. Attack dogs were brought back onto the section for the first time in six years. They had been banished from the corridors of Pollsmoor by its new, reformist head. The warders patrolling the section made no attempt to conceal their hostility— they were itching to retaliate.

Indeed, warders had in fact retaliated. In the immediate aftermath of the attack, warders reverted to an old ethos learnt under apartheid: they beat the stabbers heavily. Perhaps it happened because the head of the prison was on extended leave at the time. Their action was, on the one hand, a release of anger and an expression of fear, but it was also an attempt to get information, to find out who had ordered the assault, and why. The prison's management did discover that the assault had been planned by the 26s and 27s. But they were not sure why.

But the ndotas I spoke to were all sure. "It's obvious," I was told. "They didn't just want to stab him. They wanted to kill him. That means the 26s were attempting to raise their flag over the prison,"—meaning that it was a bid by the 26s to wrest control of Pollsmoor, both from the warders and from the 28s.

In the weeks prior to the stabbing, a series of events had upset and unsettled the inmates. For one, the head of the prison, a charismatic figure with a strong presence, had gone on extended leave. Inmates could not articulate why his absence unsettled them. Indeed, it remains something of a mystery to me. But they all insisted on underlining the importance of his absence. Second, there had been a surprise search raid in the prison, conducted in the middle of the night by warders brought in from other prisons. The raid had unearthed an armoury of homemade weapons and large stashes of mandrax and cannabis. Finally, in the weeks prior to the assault, senior warders had discovered and closed a "hole" between two sections of the prison. A young, junior warder, in charge of a notorious 26-dominated cell on D section, had been bullied into allowing senior 26 members to swap sections at night, thus creating a 26 thoroughfare through the prison. The warder's actions had been discovered. He was transferred and the hole closed.

In short, the events preceding the stabbing amounted, in inmates' eyes, to a concerted staff campaign to tighten control over the prison. Large quantities of weapons and merchandise had been confiscated, movement had been restricted. One of the goals of the stabbing was to win back space that had been forfeited.

And, indeed, the day after the stabbing, the 26s spread a rumour and ensured that it reached the ears of the staff: another senior warder—and they named him—would be stabbed within the next week. The rumour had its desired effect. The warder in question was removed from the prison and put on office duty. There was infighting among senior staff: recrimination, finger-pointing, talk of demotion for those who were not doing their jobs. In the wake of the stabbing, the staff were undoubtedly both afraid and divided. What the 26s had in mind was the creation of a prison environment lucidly described by Gresham Sykes:

The social system of the prison finally reaches a point where the inmates have established their own unofficial version of control. The custodians, in effect, have withdrawn to the walls to concentrate on their most obvious task, the prevention of escapes. The outward guise of the custodians' dominance within the walls is preserved, to be sure, for the inmates are still counted, some infractions of the rules are still punished, and prisoners continue to be marched back and forth from their cells. But surveillance has grown lax and guards are careful not to antagonise influential inmates. Institutional supplies are looted with relative ease and goods flow in freely from the outside world. Prisoners administer their own stern justice to inmates who have broken the inmate code and officials seek the advice of their captives in regard to cell and job assignments.48

If one of the goals of the stabbing was to win back control from the warders, another was to dominate the 28s. Were the 28s to acknowledge that the 26s had indeed "raised their flag" over the prison, they would have to submit to 26 authority in the day-to-day management of inmate life. The 26s would get to choose who is appointed to the "food spans"—the inmates who go to the kitchen at meal times to fetch the food trolleys. This is a crucial function, for it gives inmates the freedom to move through the prison, and thus to communicate. The 26s would also get to control much of the internal drug market: it would be understood that the 26-aligned merchants would have privileged access to drug consumers in the prison. "The 26s stab and raise their flag," an inmate told me. "Then they rule Pollsmoor; they get all the power and the respect." In other words, since the 26s were the ones who pushed the warders back to the periphery of the prison, they are the ones who reap the benefits.

In the wake of the stabbing, however, the 28s refused to acknowledge that the 26s had raised their flag. The reason was simple. The stabbers had been beaten publicly in the aftermath of the attack, and they had cried, the 28s claimed. They had squealed, naming their co-conspirators on the spot. They had shown no fortitude, no stoicism. The stabbing thus meant nothing.

Once again, the meaning of the stabbing turned on an appeal to a lost authority, to the way things were done when ndotas were ndotas. One senses that the stabbers were always going to fail the test, that every prison-gang action always fails the test, that the authority to which appeal is made is designed to be elusive. The crisis of legitimacy is in essence a crisis of masculinity. The real man, the ndota vleis en bloed, stands in the mists of the past, in the days "when the Number was pure". Everyone who tries to emulate him will fail. Nobody will ever be a man again.

It is a strange and interesting use of tradition. Usually, the past is reified and traditionalism thus established in order to confer authority upon actions in the present. Here, the function of tradition is inversed. It is used to emasculate the actions of the present, to hollow them out. The past represents something deeply painful: an impossibility.

No doubt, the past was never quite what it is made out to be. If I had conducted this research 20 years ago, it is probable that veteran ndotas would have moaned, as they do now, about the erosion of tradition, about how men were really men when they were recruited. All the testimony I gathered from the veterans is more or less idealised. Indeed, it appears that prison gangs have been inventing a pure past for themselves since the very beginning. That is probably why the Nongoloza myth has been pushed forward nearly 100 years. There had to be an original purity, a lost wholeness, right from the beginning.

Nonetheless, it would probably be a mistake to argue that nothing ever changes, that history simply repeats itself. One can only guess at the sources of legitimacy crises in prison-gang ranks 20 years ago, but there can be little doubt that they are qualitatively different from those of the present. The emergence of gangster capitalism and its multi-million rand industries in the late 1980s, and the simultaneous interpenetration of the ethos of the prison and the ethos of the street, has undoubtedly changed prison-gang culture forever. Nongoloza the robber baron and Nongoloza the prisoner of fortitude, solidarity and restraint are likely to spar with one another for a long time to come.

Chapter 5: Warders and gangs

In October 2002, when the Minister of Correctional Services came to Pollsmoor—the day Rashid was supposed to stab a fellow inmate—he gave a 45-minute speech, a few sentences of which were devoted to prison gangs. "You are here because you have done terrible things to people on the outside," he told the inmates in the audience. "Don't do terrible things to people on the inside. Don't rape people here. The gangs are unacceptable and must go."

Four years earlier, the Department of Correctional Services had given a presentation on gangs to Parliament's Correctional Services Portfolio Committee. There, the department had candidly stated that its "efforts in combating gangsterism can restrict their influence but can never eliminate the problem". The department also acknowledged that in many prisons, its staff had ceded spheres of control to the gangs, that warders who sought to close down gang space in prison were likely to be victimised. "Personnel are often intimidated by the gangs in order to get the opportunity to promote gangsterism," the department informed the portfolio committee. "Members [of the department] are eventually assaulted if it is noticed that they are the obstacles in [gangs'] way…. In the places where prisoners are actively involved in gang activities, the normal functioning of the institution is disturbed."49

Finally, in December 2003, the department released a draft white paper—a 20-year vision for correctional services. Two paragraphs of the 100-page draft are devoted to gangs, the second of which reads: "The pervasive manner in which prison gangs assert control over the management of correctional centres requires an anti-prison-gang strategy to be adopted by correctional management."50 Nowhere in the document, however, does the department suggest what such a strategy might entail.

It is clear from the three statements above both that the department is deeply concerned with the problems gangs pose for the governance of prisons, and that it is not sure what to do about them. This is nothing new. Throughout the 20th century, successive governments have oscillated wildly, not only over strategies and tactics, but also over fundamental goals. Some administrators have lobbied to outlaw prison gangs; others have attempted to emasculate them by isolating their leaders; reformers have hoped to tame them by improving harsh prison conditions; still others have attempted to win over Number-gang leaders and enlist their help in dismantling the gangs.51 The Number gangs have, of course, outlived all these administrations.

Without a coherent policy from the centre, it appears that the department's approach to gangs varies from region to region and area to area, depending on the character and predilections of senior personnel. My only substantive experience was of Pollsmoor Admission Centre, and that is the prison I discuss here.

In 1997, an enterprising reformer, Johnny Jansen, took charge of Pollsmoor Admission Centre. He inherited a prison which, by all accounts, was in a state of alarming disarray. The transition to democracy had unsettled the institution immensely. Although the prison was ostensibly subject to the rule of law, and despite the fact that corporal punishment, solitary confinement as a form of punishment and, indeed, the whole array of informal discipline had been scrapped, the relationship between warders and inmates, and between inmates themselves, was extremely violent. In 1995, two years before Jansen took office, there were 78 recorded assaults by staff members on prisoners, and 219 assaults by prisoners on one another. The staff itself were fractious and divided, primarily along racial and ethnic lines. According to one staff member who was at Pollsmoor in the mid-1990s: "We were purely in defence mode. Our only aim was to prevent members of staff from being hurt. The gangs could run the prisons as far we were concerned, as long as they didn't touch one of us. So we patrolled with dogs, we had batons on our hips. Our only concern was that they know that if they touch us they will bleed."

According to the testimony of inmates, warders had lost control of the day-to-day running of the prison. "They were too scared even to take us into the yard for exercise," an inmate who had been in Pollsmoor Admission Centre in 1997 told me. "We exercised once a month. There was no sport, no recreation—they were afraid of giving us the freedom to move around. So we sat there in our sections, day in and day out. And the sections were packed. Pollsmoor was nearly 200 per cent overcrowded. We lived on top of each other, in the heat and the cold. The Number ran the sections, not the warders."

On taking charge of the prison in 1997, Jansen began to de-escalate the decades-long war between warders and inmates. Warders were made to understand, for the first time, that the prohibition on violence was for real, that they would be dismissed for assaulting prisoners. Warders were ordered to leave their batons outside the prison gates. The dogs that had customarily patrolled the corridors of the prison were redeployed at the prison gates.

In the awaiting-trial section of the prison, Jansen separated gangsters from non-gangsters, housing them on separate floors. On the floor that housed gangsters—D floor—he identified Number-gang leaders and began talking to them. Within a year of his appointment, he had set up a committee of inmate delegates, each in charge of a different portfolio—sports, food, health, sanitation and so forth. He made a tacit pact with them. He would lift restrictions on movement—increase exercise time, begin sports and recreation again—and consult inmates in regard to aspects of the running of the prison, but only as long as gang leaders took joint responsibility for the maintenance of order in the prison. The success of this initiative hinged on Jansen's capacity to display his integrity to inmates. He had to give his personal assurance that none of his warders would deploy arbitrary violence. He had to ensure that when reasonable complaints were made—about the quality of food, or medical attention, for instance—he was able to deliver. The relationship between warders and gangs under apartheid was one of predictable modes of violence. Jansen had to replace this with a new ethos, one that consisted of predictable modes of mutual agreement.

Jansen also began collaborating with non-governmental organisations. A series of workshops was established in Pollsmoor on conflict management, life skills and personal development. The majority of participants were gang leaders. At the very least, the workshops brought a lengthy peace to the relationship between gangs.

The results of the new atmosphere in the prison were remarkable. By 2000, reported assaults by warders on prisoners were down to 11 (from 78 in 1995). Reported assaults by prisoners on one another were down to 47 (from 219 in 1995).52 Visitors to the prison expressed astonishment at the rapid and tangible changes. One non-governmental-organisation worker, who has visited Pollsmoor regularly since the mid-1990s, told me: "It was extraordinary to see the prison change before my eyes. In 1997, you walked through the corridors with armed guards. The tension was so severe you felt the goose bumps rise on your skin. By 1999, you walked freely through the prison, talking to inmates casually. The food was the same, the overcrowding was the same, but the atmosphere was that of a different planet."

The speed and degree of the de-escalation was remarkable; indeed, it flies in the face of experience in many other parts of the world. Jansen's strategy was high-risk. In a context where the material conditions of incarceration were, if anything, deteriorating, he raised inmate expectations about the quality of prison life. In a context where warders felt physically threatened, he disarmed them and began to demand that they take the idea of rehabilitation seriously. Reform efforts easily backfire in such circumstances. The raised expectations of inmates transmute into frustration and anger. The disarming of warders instils resentment and fear. The old system of control is broken down, but nothing viable replaces it. Inmates sense the uncertainty, and begin to push the boundaries of control as far as they can.53 How did the reform initiative at Pollsmoor Admission Centre avoid these unintended consequences?

Perhaps the reform initiative did not, in fact, avoid these pitfalls. Perhaps the distinctive character of Pollsmoor has only masked and delayed some of the problems that accompany projects of reform. The success of reform at Pollsmoor brings to mind a war-devastated economy that suddenly achieves high levels of gross-domestic-product growth simply by opening the factories again. Only once the rudiments of economy activity are back in place do its deep, structural problems begin to manifest themselves. Like the hypothetical post-war economy, the reform project at Pollsmoor started from a very low base. Inmates and warders were armed to the teeth and engaged in internecine conflict. Jansen disarmed and de-escalated. Perhaps his success simply lies in the fact that he brought a modicum of sanity, intelligence and a great deal of personal courage to the running of an institution that had long been crazy.

By the time I arrived in Pollsmoor in 2002, the reform process was five years old and beginning to show its age. Many of the classical problems that accompany reform in difficult environments were beginning to manifest themselves. I was struck in particular by the gap between the prison's official discourse, which emphasised the primacy of rehabilitation, and the immediate and practical concerns of its staff, which concerned control and security.

The agenda of daily staff meetings was of particular interest. Meetings would customarily begin with a prayer and a brief speech by a senior official. The prayer and speech would generally encourage staff to understand the principles of rehabilitation, to treat inmates with humanity, to understand their individual problems and concerns. In short, staff members were reminded repeatedly that their overarching goal was to shore up the human soul that resided within each inmate. "Each member a rehabilitator" was a slogan often used. Yet, when the daily meetings got down to business, discussion would usually turn to questions of control and safety: restricting movement between the sections of the prisons, taking a troublesome inmate out of a food team, plugging a "hole" that had allowed gang leaders in one section to recruit new members in another.

As has been argued repeatedly in the literature on prison management and culture, the goal of maintaining social order in the prison and that of rehabilitating inmates are bound to conflict with one another. Maintaining social order in a prison inevitably sets in motion the infantilising process Goffman speaks of. Movement must be strictly controlled. Private possessions must be constantly searched. Association between inmates must be restricted and carefully monitored. Inmates returning from court must be strip-searched. Implements that could be honed into lethal weapons, like metal spoons, must be kept under control. In short, maintaining control in the prison mortifies inmates; rehabilitation is about fostering adult autonomy. The very management of a prison is constantly at odds with its goal of rehabilitation.

At Pollsmoor Admission Centre, this tension is particularly acute, its manifestations vivid. During the period in which I conducted my research, the prison was severely overcrowded. Many of the communal cells, designed to house 18 inmates, housed 40 or 50. Needless to say, privacy was the scarcest resource in the prison. Day in and day out, inmates lived in one another's lives. Even more troubling, the imperatives of controlling an overcrowded prison meant that warders were in no position to alleviate the cramped, claustrophobic conditions in the jail. Inmates were locked in their cells between 4pm and 7am every day. Although inmates were supposed to spend an hour a day in the exercise yard, most were lucky to get an hour a week. This was not the result of malevolence on the part of staff. The prison was simply so crowded that staff could not both maintain control and give each inmate an hour of daily exercise. Inmates were forced to spend most of each day idle, in cramped quarters, because the exigencies of control did not permit otherwise. The task of treating inmates with humanity and controlling them clashed every hour of every day. Inevitably, the exigencies of the latter must win out in the end. In this context, talk of rehabilitation as a primary goal smacked of madness for many warders—it simply did not make sense in the context of the reality of their day-to-day work.

Here is a particular anecdote, but it is emblematic of a general problem in prison management. During the period of my research at Pollsmoor, the 28s split into two factions. Faction A was led by an incumbent leader, whom I shall call "Buttons", Faction B by a usurper, whom I shall call "Hassan". Hassan wanted to build an army. He told his followers that whoever stabbed an officer in Faction A would get his victim's post in the gang. And whoever stabbed Buttons himself would become Hassan's right-hand man. I discovered this in the course of speaking to 28s.

In the midst of this conflict, I sat in on a meeting between Hassan and a senior warder. The warder lectured Hassan a long time, appealing to him to de-escalate the conflict in the 28s, appealing to his humanity, asking him to help make Pollsmoor a tolerable place. Throughout the interview, I sensed that the warder had insufficient information, that he was not aware that Hassan was in the thick of a power struggle, that the stakes were far too high for him to climb down now. I sensed that the warder was simply going through the motions of talking to Hassan, and that he actually did not know what to say to him. When the meeting was over, I told the warder what I knew—about the conflict between Hassan and Buttons, and about Hassan offering his soldiers senior positions if they won the war on his behalf.

"I know all that," the warder replied. "But I dare not tell Cups that I know what he's up to. If I confront him he will target me, I'll have to watch my back. I can't confront him because I don't have a plan. I can tell him I know what he's doing, but then what do I do next?"

This is a classic scenario—one in which a warder has ceded a sphere of power to a prison gang, but goes through the motions of pretending that he is still in control. The warder in question was not a bad warder: on the contrary, he was one of the most thoughtful and innovative staff members I met at Pollsmoor. The problem is that the official discourse of the prison did not guide him in how to do his job. The official discourse spoke of rehabilitation. It was entirely irrelevant to his most urgent daily task—maintaining control.

This is not to argue that prison administrators should abandon the aim of rehabilitation, still less the aim of treating inmates with humanity. But it does strongly suggest that, given current conditions in South African prisons, administrators would do well to acknowledge that the better part of their job does inevitably infantilise inmates, that inmates will keep adjusting to this infantilisation by immersing themselves in gang activity, and that a daily prerogative of prison governance will always remain the maintenance of control. The improvement of poor conditions cannot happen without control. Nor can successful rehabilitation programmes. The real question is how to recover control in a manner that frees up the space for better conditions and for rehabilitation.

The comments above might be restricted to Pollsmoor Admission Centre, but they are germane to the general question of the future of prison policy—particularly policy in relation to prison gangs.

Number gangs are here to stay in Western Cape prisons. They have survived and flourished over many decades and there is no prospect of them wilting now. For one, their newfound iconic status outside prison, at the heart of the Western Cape's underground economy, is powerful and is set to last. The idea of Nongoloza the millionaire robber baron organises the collective imagination of the underworld and animates the players in the illicit economy. The legend of prison gangs is too entrenched to disappear in the near future.

Conditions within prisons themselves also foster the longevity of the Number gangs. The recent introduction of mandatory minimum sentencing means that the prisons are increasingly filled with long-term inmates. Much of their adult lives are lived in total institutions; the wells of psychological need from which prison gangs draw run deep. Finally, the extent of overcrowding in South African prisons means that inmate experience will remain an intensely communal and infantilising one for the foreseeable future. These are precisely the conditions in which gangs flourish.

The Draft White Paper on Corrections in South Africa published in December 2003 envisages a gradual alleviation of the harsh conditions in South African prisons. It envisages a prison environment in 20 years' time in which each prisoner has an adequate sphere of privacy, the opportunity to educate him or herself and to work, and has reflected meaningfully on his or her crime. There can be little doubt that as prison staff members move towards implementing these goals, they will come up against the rival goal of maintaining order. Will it be possible to give inmates the freedom of movement and association required for meaningful work and recreation while maintaining order at the same time? How does one both grant these freedoms and contain the power of the Number gangs? Any warder in a South African prison knows from his or her daily experience that this question is paramount.

This is not a reactionary argument for a return to discipline at its most uncompromising. It is simply an appeal that the department recognises the most fundamental characteristic of the institutions it administers: they keep human beings confined against their will. Only once the department acknowledges that the maintenance of order is central will it be able to deal with the question of order in a way that makes rehabilitation possible.

In this regard, there are important lessons to be learnt from a particular slice of the Pollsmoor experience. As mentioned earlier, one of Jansen's first moves as head of Pollsmoor Admission Centre was to establish a committee of inmate delegates, each member of the committee assigned to a particular portfolio—food, recreation, health, and so forth. The 26s, 27s and 28s were not represented as distinct blocks on the committee. But Jansen did ensure that each committee member was a prominent gang leader.

The nature of his interaction with the committee was akin to a pact. He increased spheres of freedom in the prison—freedom of movement in particular. In exchange, the gangs were forbidden to use new freedoms malevolently. An example of such a tacit pact might be, for instance: "exercise time will be doubled, I will permit soccer and cricket matches in the yard, but the moment there is an assault in the yard, recreation is suspended". Or: "I will set up channels for you to complain about the quality of the food, and I will do my best to address your dietary complaints, but you must play by my rules—no stabbings, no assaults". The running of the prison bureaucracy loses its dictatorial arbitrariness; managerial decisions are discussed with inmates and become predictable and rational. But, in exchange, inmates must observe a prohibition on ritual gang violence.

It is a simple quid pro quo. Conditions improve. Inmates themselves assist in the running of new spheres of freedom. The prison staff allows inmates to regulate domains of prison life, but only if those domains are run according to staff rules. The moment there is a transgression, the domains of freedom shrink.

Such a strategy is, of course, laden with risks, and the establishment of inmate committees composed of prison-gang leaders has had mixed success elsewhere.54 The fabric of the pact will always be fragile, since a degree of animosity is built into the very structure of the relationship. Inmates are bound to resent the pact. They will be asked, in essence, to collaborate in maintaining the rules of their custodians, and they will inevitably strain against the roles they are being offered. The pact is sure to fail whenever the new domains of freedom offered to inmates are merely nominal. In that case, there is no quid pro quo, merely an opportunity for gang leaders to push the staff to the margins of the prison. The pact is also sure to fail when the prison administration's behaviour becomes unpredictable. Arbitrary violence, the withdrawal of privileges for reasons that are unintelligible to inmates—these would destroy the pact instantly.

But when the relationship works, the politics of prison life becomes increasingly benign. Gangs remain, and go through the motions of practising their rituals. But the pernicious aspects of gang activity, like the taking of blood at initiation and promotion, begin to wither. The administration would have shored up order, not through a simple show of force, but through the establishment of a predictable and demonstrably rational regime, one in which inmates benefit when they play by the rules.

The goal of such policy is not to eradicate gangs. It is to minimise violence and to transfer spheres of social control from inmates to staff. A policy to eliminate gangs will achieve neither of these goals.

The White Paper insists that the primary goal of the Department of Correctional Services is to rehabilitate. Yet it gives no policy directives in regard to the maintenance and management of overcrowded prisons. It mentions that warders have ceded control to gangs in many prisons, but gives no directives as to how warders can win it back. It is almost as if the authors of the White Paper regard the heart and soul of a warder's job as "dirty work", work that is best not spoken about.

Work that is not spoken about becomes work done poorly. As the disjuncture between head office policy and the reality of prison life grows, so warders gradually sidestep policy and do their work according to the dictates of their own informal "rules"—"rules" that involve both violence and corruption. The result is that the White Paper's complaint about the poor quality of staff could become a self-fulfilling prophecy. Moreover, the White Paper's own primary goal, rehabilitation, could be undermined: the best rehabilitation programmes in the world will be ineffectual and meaningless if the moral and physical fabric of prison life is at odds with the values and precepts of rehabilitation.


1 For Nongoloza's life testimony, see Department of Justice, Annual Report for the Year 1912, 237–240. For an extraordinary biography of Nongoloza, see Charles Van Onselen, The Small Matter of a Horse (Johannesburg: Ravan, 1984). For a brief history of the Ninevites, the band of outlaws Nongoloza founded, see Charles Van Onselen, New Babylon and New Nineveh: Everyday life on the Witwatersrand 1886–1914 (Johannesburg: Jonathan Ball, 2001) 368–397.

2 On South African prison gangs in the late 20th century, see Nicholas Haysom, "Towards an Understanding of Prison Gangs" (Institute for Criminology, University of Cape Town, 1982); Breyten Breytenbach, The True Confessions of an Albino Terrorist (London: Faber & Faber, 1984) 272–279; Rodney Keswa, "Outlaw Communities" (unpublished manuscript, 1975); J.M. Lötter and W.J. Schurink, Gevangenisbende: 'n Ondersoek met spesiale verwysing na nommerbendes onder Kleurlinggevangenes (Pretoria: RGN [Verslag S115], 1986); J.M. Lötter, "Prison Gangs in South Africa: A description", South African Journal of Sociology, 19(2) (1989), 67–75; W.J. Schurink, "The World of the Wetslaners: an analysis of some organisational features in South African Prisons", Acta Criminologica, vol. 2, no. 2 (1989), 60–70; W.J. Schurink, E. Schurink, and J.M. Lötter, "Numbers gangs in South African prisons: an organisational perspective", paper delivered to Association for Sociology in Southern Africa (July 1986); Eduard Fagan, "S v Nero en 2 Ander", Stet, (3)1 (February 1985), 22–26; Sasha Gear and Kindiza Ngubeni, Daai Ding: Sex, Sexual Violence and Coercion in Men's Prisons (Johannesburg: Centre for the Study of Violence and Reconciliation, 2003).

3 See Dirk Van Zyl Smit, "Change and Continuity in South African Prisons", in Robert Weiss and Nigel South (eds), Comparing Prison Systems: Toward a Comparative and Internal Penology (Sydney: Gordon and Breach Publishers, 1998) 401–426; Dirk Van Zyl Smit, "Prisoners' Rights", South African Human Rights Yearbook, vol. 5 (Durban: Centre for Socio-Legal Studies, 1995) 268–280; Dirk Van Zyl Smit, South African Prison Law and Practice (Durban: Butterworths, 1992).

4 Haysom, "Towards An Understanding of Prison Gangs".

5 Lötter and Schurink, Gevangenisbende.

6 Department of Justice, Annual Report for the Year 1912, 37.

7 See Keswa, "Outlaw Communities"; University of the Witwatersrand Oral History Project collection, "Interview with Mankailang Karia Molokoe" (10 February 1982).

8 University of the Witwatersrand Oral History Project collection, "Interview with Sinyu Mbuli" (15 February 1982).

9 "Even the poor can be terrible" is a phrase taken from Eric Hobsbawm's description of a genre of bandit he calls "the avengers": "They are not so much men who right wrongs," Hobsbawm tells us, "but … the exerters of power; their appeal is not that of the agents of justice, but of men who prove that even the poor and weak can be terrible." See Hobsbawm, Bandits (2nd edn., New York: Pantheon Books, 1981), 58.

10 For a history of the Ninevites, see Van Onselen, New Babylon, New Nineveh, 368–397.

11 See Anthony K. Appiah, "Into the Woods", New York Review of Books (18 December 2003), 46–51, for a discussion on the relationship between the meaning of stories and the manner in which they are told.

12 Mary Douglas, Purity and Danger (London & New York: Routledge, 2002), 77–79.

13 The term "total institution" was coined by Erving Goffman. He defines it thus: "First, all aspects of life are conducted in the same place and under the same single authority. Second, each phase of the member's daily activity is carried on in the immediate company of a large batch of others, all of whom are treated alike and required to do the same thing together. Third, all phases of the day's activities are tightly scheduled, with one activity leading at a prearranged time to the next, the whole sequence of activities being imposed from above by a system of explicit formal rulings and a body of officials. Finally, the various enforced activities are brought together in a single rational plan purportedly designed to fulfil the official aims of the institution." Goffman lists five categories of total institutions: institutions of care, such as mental hospitals; institutions of quarantine, such as leprosaria; jails, POW camps and concentration camps; institutions of work such as army barracks, ships and boarding schools; and, finally, cloisters such as abbeys, convents and monasteries. See Goffman, Asylums: Essays on the Social Situation of Mental Patients and Other Inmates (London: Penguin, 1961), 16–17.

14 Goffman, Asylums, 24.

15 Goffman, Asylums, 68.

16 Gresham M. Sykes, The Society of Captives: A Study of a Maximum Security Prison (Princeton: Princeton University Press, 1958), 79.

17 For a powerful argument that the exercise of power in prisons can, in fact, facilitate a degree of autonomy among inmates, see P. Young, "The Concept of Social Control and its Relevance to the Prisons Debate", in A. Bottoms and R. Light (eds), Problems of Long-Term Imprisonment (Aldershot: Gower, 2001), 97–114.

18 Sykes, The Society of Captives, 21.

19 Sykes, The Society of Captives, 82.

20 Goffman, Asylums, 60.

21 Goffman, Asylums, 41 (emphasis added).

22 See Goffman, Asylums, 61–64.

23 See Chapter 2 for an extended discussion of a "carry-on".

24 This characteristic of the Number gangs would not have surprised Goffman. "Sustained rejection of a total institution," he writes, "often requires sustained attention to its social organization, and hence, paradoxically, a deep kind of involvement in the establishment." Goffman, Asylums, 62.

25 See, inter alia, Hugh Lewin, Bandiet: Seven Years in a South African Prison (Cape Town & Johannesburg: David Philip, 1981); Breytenbach, The True Confessions of an Albino Terrorist.

26 Author's interview with Vincent Shabangu (Macasser, 11 June 2003).

27 Author's interview with Jeremy Veary (Cape Town, 30 September 2002).

28 See Kriegler Commission, Commission of Inquiry into Unrest in Prisons Appointed by the President on 27 June 1994 (Pretoria: February 1995).

29 There has been a great deal of debate in the prisons literature about whether inmate subculture is animated by the institutional logic of the prison or by the cultural and political identities inmates bring in with them from the outside. For a useful discussion on this debate, see R. Sparks, A. Bottoms and W. Hay, Prisons and the Problem of Order (Oxford: Clarendon Press, 1996) Chapter 1. It should be clear that the discussion above attempts to dissolve this distinction. Number gang ideology and practice under apartheid is unintelligible unless one grasps that its members are constituents of a disenfranchised majority and understand themselves thus. Yet their practice—and the dilemmas that shape their practice—is irreducibly an inmate practice, moulded by the conditions of the total institution. In other words, the Number gangs are overdetermined by prison and society.

30 "Nangampela" is an isiZulu word that means, literally, "There it is, indeed!"

31 Agter die berge (behind the mountains). It means a period of reflection. Po went agter die berge to think about how to fight the whites; it is where he invented his secret language.

32 Most of these labour stations were built in the 1950s. District farmers' associations were permitted to build "prison farm outstations", which were then managed by the Department of Prisons. Farmers could employ convicts in proportion to their contribution to the construction of the prison. See T.M. Corry, Prison Labour in South Africa (Cape Town: Nicro, 1977).

33 Gore Vidal, Palimpsest: A Memoir (London: Abacus, 1996), 101–102.

34 Everyone who has written on prison gangs, without exception, assumes it to be true. See Keswa, "Outlaw Communities"; Haysom, "Towards an Understanding of Prison Gangs"; and Lötter and Schurink, Gevangenisbende.

35 This is a pseudonym.

36 See Dirk Van Zyl Smit, South African Prison Law and Practice, 33-43.

37 26s could also take blood, but, generally speaking, only for two reasons. First, a 26 would stab in order to earn "promotion" into the 27s; second, 26s would stab in wars against the 28s, in order to "raise their flag" in the prison. I discuss the raising of the flag in Chapter 5.

38 See, inter alia, Van Dam Committee, Report of the Committee of Inquiry into Events at the Barberton Maximum Security Prison Barberton on 20 and 30 September 1983 and Matters (May 1984); Keswa, "Outlaw Communities".

39 The Number gangs distinguish between somebody who occupies an office "vleis en bloed" ("flesh and blood") and somebody who occupies an office "just for the minute". The latter has not been through the requisite initiation into his office. He is appointed to the office for a limited period and function. For example, if the Twelve Points convenes to try a case and there is no landdros in the prison, a junior silver is made a landdros "just for the minute". In the instance above, what is meant is that silver officers were given positions in the gold line without having to take blood.

40 There is some evidence that in the Ninevites' latter years (the 28s, recall, are the direct descendants of Nongoloza's Ninevites), recruits were initiated into the gang via sex. According to testimony in a 1915 murder case, "new boys" at a mining compound on the Johannesburg East Rand had to sleep with a khehla in order to join the gang. "Khehla" is an isiZulu word from Shaka's times. Among its many meanings, it refers to a man, usually a demobilised warrior, who has been given permission by the king to marry. Interestingly, the court witness testified that both senior gold and senior silver officers were "amakhehla", and could thus sleep with "new boys". See Rex vs Mkosi Mkemeseni and 16 Others (Supreme Court of South Africa, Witwatersrand Local Division, 17 August 1915).

41 Some interviews point to this syndrome—the recruitment of faux 28s on the streets—as the reason why the 28s began initiating new members with sex in the late 1980s. "People came into the 28s," one interviewee told me, "but not in the proper way. They knew nothing about the Number. We wanted to teach them a lesson, and we wanted to use their own ignorance in doing so. So we said to them: 'You know, mos, what you have learnt is wrong. To be a real 28, you have to sleep with me.' And we'd make it real by telling them in prison language, using the whole story of Nongoloza."

42 Indeed, prison gang language does not distinguish between street gangs and non-Number prison gangs. They are referred to collectively as "the fourth camp", in others words, not the 26s, 27s, or 28s, and therefore illegitimate. Their tattoos are called "vuil papiere" ("dirty papers"). Veteran ndotas who complain of the interpenetration of street and prison gangs talk disparagingly of people who "bring the fourth camp into the four corners [of the prison]".

43 See, for instance, Wilfried Schärf, "The Impact of Liquor on the Working Class: The Implications for the Structure of the Liquor Industry and the Role of the State in this Regard", master's thesis (Faculty of Social Sciences, University of Cape Town, 1984); Irvin Kinnes, From Urban Street Gangs to Criminal Empires: the changing face of gangs in the Western Cape (Pretoria: Institute for Security Studies, 2000); Don Pinnock, "State Control and Street Gangs in Cape Town: Towards an Understanding of Spatial and Social development", master's thesis (University of Cape Town, 1982).

44 See Ted Leggett, Rainbow Vice: The Drugs and Sex Industries in the New South Africa (Cape Town, London & New York: David Philip & Zed Books, 2001).

45 Los Angeles and Chicago are also characterised by the interpenetration of prison and street gangs. See, inter alia, James B. Jacobs, Stateville: The Penitentiary in Mass Society (Chicago & London: University of Chicago Press, 1977), Chapter 6; James B. Jacobs, "Street Gangs Behind Bars", Social Problems, 21(4) (1974); James B. Jacobs, "Stratification and Conflict Among Prison Inmates", Journal of Criminal Law and Criminology, 66(4) (1975).

46 I also conducted lengthy interviews with two 26 veterans at Pollsmoor. Their position in the 26s mirrored that of the veterans in the 28s exactly. In addition, I conducted about 30 interviews with Number gang veterans currently on the outside.

47 The price of drugs in Pollsmoor fluctuates but, in general, the prison value of mandrax and crack cocaine is about double its street value. The reasons for this are: (1) Smuggling drugs into prison is risky and expensive (the bribing of warders, the risk of detection). (2) The profits of the prison drug market are spread wide. Merchants sell with the permission of a gang; the gang itself takes a healthy cut and pools its income collectively. Profit margins must thus be large. (3) Drugs in prison are scarce—demand exceeds supply by a wide margin.

48 Sykes, The Society of Captives, 127–128.

49 Department of Correctional Services briefing to Parliament's Correctional Services Portfolio Committee, 27 May, 1998.

50 Department of Correctional Services, Draft White Paper on Corrections in South Africa (December 2003), 83.

51 See Van Onselen, The Small Matter of a Horse; Haysom, "Towards an Understanding of Prison Gangs".

52 See Chris Giffard, "The Prisons Transformation Project: A Perspective on the Way Forward" (Cape Town: Centre for Conflict Resolution, 2001).

53 See, in particular, Jacobs, Stateville.

54 See, in particular, Sykes, The Society of Captives; Jacobs, Stateville.


Appiah, Anthony K., "Into the Woods", New York Review of Books (18 December 2003), 46–51.

Breytenbach, Breyten, The True Confessions of an Albino Terrorist (London: Faber and Faber, 1984).

Corry, T.M., Prison Labour in South Africa (Cape Town: Nicro, 1977).

Department of Correctional Services, Draft White Paper on Corrections in South Africa (December 2003).

Department of Justice, Annual Report for the Year 1910.

Department of Justice, Annual Report for the Year 1911.

Department of Justice, Annual Report for the Year 1912.

Douglas, Mary, Purity and Danger (London & New York: Routledge, 2002) (first published in 1966).

Fagan, Eduard, "S v Nero en 2 Ander", Stet, (3)1 (February 1985), 22–26.

Gear, Sasha and Ngubeni, Kindiza, Daai Ding: Sex, Sexual Violence and Coercion in Men's Prisons, (Johannesburg: Centre for the Study of Violence and Reconciliation, 2002).

Giffard, Chris, "The Prisons Transformation Project in Western Cape Prisons" (seminar presented at the Centre for the Study of Violence and Reconciliation, Johannesburg, October 2001).

Giffard, Chris, "Out of Step? The Transformation Process in the South African Department of Correctional Services", dissertation submitted in partial fulfilment of the degree of Master of Science in Criminal Justice Studies (Scarman Centre for the Study of Public Order, University of Leicester, 1997).

Goffman, Erving, Asylums: Essays on the Social Situation of Mental Patients and Other Inmates (London: Penguin, 1973) (first published in 1961).

Grann, David, "The Brand: How the Aryan Brotherhood became the most murderous prison gang in America", in The New Yorker (16 & 23 February 2004).

Haysom, Nicholas, "Towards an Understanding of Prison Gangs" (Institute of Criminology, University of Cape Town, 1981).

Hobsbawm, Eric, Bandits (2nd edn, New York: Pantheon Books, 1981).

Jacobs, James B., Stateville: The Penitentiary in Mass Society (Chicago & London: Chicago University Press, 1977).

Jacobs, James B., "Stratification and Conflict Among Prison Inmates", Journal of Criminal Law and Criminology, 66(4) (1975).

Jacobs, James B., "Street Gangs Behind Bars", Social Problems, 21(4) (1974).

Keswa, Rodney, "Outlaw communities" (unpublished manuscript, 1975).

Keswa, Rodney, "King Nongoloza Made Men of Women", Sunday Times Extra (26 September 1976).

Keswa, Rodney, "The King Recruits New Gang in Jail", Sunday Times Extra (3 October 1976).

Keswa, Rodney, "I Saw a Jail Gang Execution", Sunday Times Extra (17 October 1976).

Keswa, Rodney, "Secret Salutes Ex-Cons Give a Jail Boss", Sunday Times Extra (24 October 1976).

Keswa, Rodney, "Rival Spoilers Were Recruited in Prison", Sunday Times Extra (7 November 1976).

Kinnes, Irvin, From Urban Street Gangs to Criminal Empires: the changing face of gangs in the Western Cape (Pretoria: Institute for Security Studies, 2000).

Kriegler Commission, Commission of Inquiry into Unrest in Prisons Appointed by the President on 27 June 1994 (Pretoria, February 1995).

Leggett, Ted, Rainbow Vice: The Drugs and Sex Industries in the New South Africa (Cape Town: David Philip, 2001).

Lewin, Hugh, Bandiet: Seven Years in a South African Prison (Cape Town & Johannesburg: David Philip, 1981).

Lötter, J.M. & Schurink, W.J., Gevangenisbendes: 'n Ondersoek met spesiale verwysing na nommerbendes onder Kleurlinggevangenes (Pretoria: RGN [Verslag S115], 1986).

Lötter, J.M., "Prison Gangs in South Africa: A description", South African Journal of Sociology, 19(2) (1989), 67–75.

Mathiesen, Thomas, The Defences of the Weak (London: Tavistock, 1965).

Morgan, Rod, "Imprisonment: Current Concerns and a Brief History since 1945", in Maguire, Mike, Morgan, Rod and Reiner, Robert (eds), The Oxford Handbook of Criminology (Oxford: Clarendon Press, 1997).

Pinnock, Don, "State Control and Street Gangs in Cape Town: towards an understanding of spatial and social development", master's thesis (University of Cape Town, 1982).

Schärf, Wilfried, "The Impact of Liquor on the Working Class: The Implications for the Structure of the Liquor Industry and the Role of the State in this Regard", master's thesis (Faculty of Social Sciences, University of Cape Town, 1984).

Schurink, W.J., "The World of the Wetslaners: an analysis of some organisational features in South African Prisons," Acta Criminologica vol. 2, no. 2 (1989).

Schurink, W.J., Schurink, E. and Lötter, J.M., "Numbers gangs in South African prisons: an organisational perspective", paper delivered to the seventeenth Annual Society Congress, University of Natal (1986).

Sparks, Richard, Bottoms, Anthony and Hay, Will, Prisons and the Problem of Order (Oxford: Clarendon Press, 1996).

Sykes, Gresham M., The Society of Captives: A Study of a Maximum Security Prison (Princeton: Princeton University Press, 1958).

Van Dam Committee, Report of the Committee of Inquiry into Events at the Barberton Maximum Security Prison Barberton on 20 and 30 September 1983 and Related Matters (May 1984).

Van Onselen, Charles, The Small Matter of a Horse: The Life of "Nongoloza" Mathebula, 1867–1948 (Johannesburg: Ravan, 1984).

Van Onselen, Charles, New Babylon, New Nineveh: Everyday Life on the Witwatersrand (Johannesburg: Jonathan Ball, 2001) (originally published in two volumes in 1982).

Van Zyl Smit, Dirk, "Change and Continuity in South African Prisons", in Weiss, Robert and South, Nigel (eds), Comparing Prison Systems: Toward a Comparative and Internal Penology (Sydney: Gordon and Breach Publishers, 1998), 401–426.

Van Zyl Smit, Dirk, "Prisoners' Rights", South African Human Rights Yearbook, vol. 5 (Durban: Centre for Socio-Legal Studies, 1995), 268–280.

Van Zyl Smit, Dirk, South African Prison Law and Practice (Durban: Butterworths, 1992).

Van Zyl Smit, Dirk, "Adopting and Adapting Criminological Ideas: Criminology and Afrikaner Nationalism in South Africa", Contemporary Crises, 13 (1989), 227–251.

Vidal, Gore, Palimpsest: A Memoir (London: Abacus, 1996).

Young, P., "The Concept of Social Control and its Relevance to the Prisons Debate", in Bottoms, A. and Light, R. (eds), Problems of Long-Term Imprisonment, (Aldershot: Gower, 2001).

© Centre for the Study of Violence and Reconciliation

Home | Programmes | Research | Consultancy | Coming Events |
Staff | Internships | Annual Report | Background | Aims |
African Transitional Justice Research Network |
Justice in Perspective | Links

eXTReMe Tracker