Themba Anton Sukude was convicted in April 2006 of murdering four men in Newcastle, a town in northern KwaZulu-Natal, over eleven months in 2004 and early 2005. The case is remembered less for the man than for the way it was proved. For the first two killings the state had eyewitnesses: two women who survived the attacks and identified him. For the last two it had no witness, no confession that a court could use and no physical trace of him. Judge Vivienne Niles-Duner of the Pietermaritzburg High Court convicted him of those killings too, on the strength of the similarities between all four attacks and on the evidence of a police psychologist, Dr. Gerard Labuschagne, who had carried out what is known as a linkage analysis.
The South African Press Association and The Mercury reported on the day of sentencing that this was the first time a South African court had convicted a serial offender of crimes linked to him by such an analysis. Labuschagne's own later account of the case is more careful: he lists it as one of several South African trials in which linkage evidence was admitted as support for similar fact evidence. Either way, the case became a reference point in the research literature on crime linkage, and British researchers later built a mock trial around it to test how ordinary jurors would treat the same kind of evidence.
Sukude was sentenced on the day of his conviction to five life terms and further terms of imprisonment, to run concurrently, and the judge directed that he not be considered for parole until he had served at least 25 years.
The High Court judgment gives Sukude's date of birth as May 16, 1974. The press reported his age as 31 at the time of his conviction. The few personal details on record come from the plea in mitigation his counsel put to the court before sentence. He had gone as far as grade five at school. Before his arrest he had been working and earning about R40 a day. He had two children. He had been in custody since his arrest in March 2005.
He also had a criminal record. The judge described it as "an impressive list of previous convictions" going back to 1990, and noted that most of the convictions were more than ten years old. The judgment does not list the offenses. From that record the judge drew the conclusion that he had no respect for the law or for the rights of other people.
At the time of the attacks he was living in a tent in Newcastle. He denied in court that it was his.
The court records give the names of three of the four men. The fourth was never identified, according to the judgment.
Veli Ncamo was killed in February 2004. Little is on record about his age, his work or his family. A woman who was with him survived and testified at the trial.
Tebogo Mthethwa was killed on October 27, 2004. He was a minibus taxi driver. A woman who was with him also survived and testified.
The third man was found dead on November 26, 2004. The judgment refers to him only as unidentified, and his name did not appear in the reports of the trial. No woman came forward as a witness to that attack.
Dhanpalan Govender was found on January 8, 2005, critically injured. He was taken to hospital and died, without regaining consciousness, about two days after the attack, according to Labuschagne's account. He could not tell anyone what had happened. Money he had with him was not taken, a detail the court and the psychologist both treated as part of a pattern.
All four men died of head injuries.
The four attacks took place over almost exactly eleven months, all of them at night and all in Newcastle. The first came in February 2004 and the second in October. In each of those two cases a man was killed and a woman who was with him survived and went to the police.
The police connected those two cases quickly. Labuschagne later wrote that the investigating officer contacted him shortly before the arrest, and that the investigation was then already focused on a single offender.
The November and January killings were different in one respect: no surviving witness came forward in either case.
The break in the case came from a chance conversation.
On or about March 8, 2005, a 68-year-old man who earned his living collecting scrap metal for resale met Sukude and fell into conversation with him. According to the man's evidence at trial, Sukude said things in that conversation that connected him to the attacks.
The older man did not let it go. With another man he went looking for Sukude. The second man telephoned the police, and when officers arrived the witness pointed Sukude out. Inspector Henry Mtshali of the Newcastle police made the arrest. The judgment gives the arrest date as March 8, 2005, in most places and as March 9 in one.
After the arrest Sukude pointed out to police a tent as the place where he lived.
Labuschagne wrote later that during questioning Sukude admitted the first two murders to police, but refused to take part in a pointing-out of the crime scenes or to make a formal confession. Under South African law, he explained, an admission of that kind made to a police officer who is not a commissioned officer cannot be used as evidence. What he had said to the man he met was different: a statement to a member of the public could be put before the court.
Two procedures followed. On March 10, 2005, one of the women was shown seven photographs by Inspector Oupa Vilakazi and picked out the sixth, which was of Sukude. The court found that Vilakazi himself did not know who the suspect was, so he could not have steered her choice. On March 16 the other woman attended an identification parade run by Inspector Ben Viviers of the Newcastle crime prevention unit. She looked along the line for about four seconds and pointed to Sukude.
The defense made much of one fact about that parade: Sukude was the only man in it wearing leg irons. The prosecution's answer was that Viviers had covered the irons, and photographs taken at the parade by Inspector Jakobus Olivier showed that they were not visible from where the witness stood. The court accepted that she had recognized his face, "and not because she could see leg irons on him."
For the November and January killings the prosecution had neither a survivor nor an eyewitness, and nothing physical tied Sukude to either scene. Labuschagne, a forensic psychologist and senior superintendent in the South African Police Service, who worked in the police's investigative psychology unit within the Serious and Violent Crimes head office, had been brought in by the investigating officer, Inspector Theunis du Plooy. After the arrest he met du Plooy again, and he recommended that all four murders be charged. He offered to give evidence. The decision to charge the last two murders on the basis of similar fact evidence, supported by a linkage analysis, was taken by the prosecution in consultation with the Director of Public Prosecutions, according to his account.
Linkage analysis asks a narrower question than a criminal profile. It does not try to say who committed a crime. It asks whether a set of crimes was committed by the same person, by comparing what the offender did in each. Labuschagne stressed that distinction in his account of the case: he did not interview Sukude, because the aim of his report was to give an opinion on whether one offender was responsible for the crimes, not on who that offender was.
His sources were the investigating officer, visits to the scenes after the event, the police case dockets with their photographs, sworn statements and post-mortem reports, and an interview with one of the women who survived. He followed a step-by-step method published by the researchers Hazelwood and Warren in 2003, and wrote his report under five headings: aim, information sources, key terms, linkage analysis and conclusion.
The analysis compared the four attacks on victim type, time, place, cause of death and what was or was not taken. Labuschagne found a consistent pattern across all four, and he called one element of it the offender's signature: the failure to take valuables that were there to take. He told the court that this was "tantamount to Sukude leaving his calling card," as The Mercury reported.
He also explained to the court, and later to reporters, some of the general ideas behind the method. Serial offenders, he said, do not usually stop until they are arrested, even when police are active in the area. Their behavior can change with experience, intelligence and motivation. Repeating their crimes gives them a sense of familiarity and control, and they tend to keep to "comfort zones," staying in or near the same area. He described linkage analysis as "a powerful tool" for detecting and prosecuting serial offenders.
The case was heard in the High Court of South Africa, Natal Provincial Division, sitting in Pietermaritzburg, under case number CC 34/06, before Judge Vivienne Niles-Duner. There was no jury; South Africa does not use juries, and the judge decided both the verdict and the sentence. An interpreter was used during the proceedings.
Sukude faced nine counts: four of murder, with other charges (including two counts of robbery with aggravating circumstances and one of malicious injury to property). Under section 220 of the Criminal Procedure Act he made formal admissions which, together with undisputed state evidence, established that all nine offenses had been committed. The only question for the court was whether he was the man who committed them.
The state's witnesses were the man who reported the conversation, the two surviving women, several police officers who had attended the scenes or run the identification procedures, the investigating officer and Labuschagne. The defense did not attack the honesty of the state witnesses. Its case was that the women were mistaken.
Sukude gave evidence. He denied any knowledge of the attacks and said he had never spoken to the man who reported the conversation before his arrest. He said that on the day in question he had been walking into town from a job. He denied that the tent he had pointed out was his home: according to him, the police had asked where another man lived, and he had shown them that man's place. He suggested that the witnesses and the investigating officer had conspired against him, and that the woman at the parade had picked him out only because of his leg irons.
The judge did not believe him. She described him as "clearly not unintelligent" but "unimpressive," a witness who was "making up his story as he went along." His account of showing the police the tent was "highly improbable," and his version as a whole was not reasonably possibly true.
The judgment was delivered on April 25, 2006. The judge treated the man who reported the conversation and each of the two women as single witnesses on the points that mattered, which in South African law calls for caution. Applying that caution, she found that their evidence could safely be relied on. His account of the conversation was supported by the women's evidence. Police evidence supported the women's statements that there was enough light at each scene to see the attacker's face, and the court found that each woman had a proper opportunity to see him. On that evidence Sukude was convicted of the February and October murders and other charges.
The November and January killings depended on inference. The court relied on the earlier South African decision S v D, reported at 1991 (2) SACR 543, on similar fact evidence. It listed the features the four attacks shared, among them fatal head injuries, attacks at night, the timing of the later attacks, within one to two and a half months of the October attack, and their location in the same part of Newcastle. From all of this the judge found that "the irresistible inference" was that the same person had committed all four crimes. Labuschagne's linkage analysis, she noted, had reached the same conclusion.
He was found guilty on all nine counts.
The judge sentenced him the same day. She called the crimes "savage and brutal" and told him that his murders "were committed in a way which revolts in the extreme." "You have shown no respect for human life," she told him. She described him as a menace to society, found that his prospects of rehabilitation were "zero," and said that he should be taken out of society permanently. She said she hoped he would stay in prison for the rest of his life.
Defense counsel made no submissions on substantial and compelling circumstances beyond his personal details. The sentences, as reported by The Mercury and recorded by Labuschagne, were:
life imprisonment on each of the four murder counts, on the grounds that the killings were premeditated or were committed in the course of robbery with aggravating circumstances;
a further life term and a ten-year term on two of the other charges;
15 years on each of the two counts of robbery with aggravating circumstances;
12 months for malicious injury to property.
The sentences run concurrently, according to Labuschagne. Under section 276B of the Criminal Procedure Act the judge directed that Sukude "shall not be considered for parole until he has served at least twenty five (25) years."
No appeal against the convictions or the sentences was reported.
The Sapa report of April 25, 2006, carried by IOL, began by calling Sukude the first person in South Africa convicted of crimes attributed to him solely through a psychologist's linkage analysis. Bob Frean's report in The Mercury the next day went further, describing the judgment as the first time in South African legal history that psychological linkage analysis had been accepted to convict a serial murderer, and presenting it as a new tool for investigators and prosecutors.
The two reports did not agree on how the counts divided. Sapa said he was convicted of two murders and other offenses on the witnesses' evidence, and of two more murders and an aggravated robbery for which there was no conventional evidence, bringing the total to six crimes. The Mercury said he was convicted of six crimes on the witnesses' evidence and of two more murders and an aggravated robbery solely on the strength of the linkage analysis. The court itself convicted him on nine counts.
The word "solely" also oversimplifies what the judge did. Her judgment rests the November and January convictions on the circumstantial features the four attacks had in common, judged by the rules on similar fact evidence, and treats Labuschagne's analysis as reaching "the same" conclusion. In his own chapter on the case, published in 2012, Labuschagne describes linkage reports in South Africa as admitted in support of similar fact evidence, and names the Sukude case as one example among several, not as the first.
Labuschagne published an account of the analysis in 2006 in the Journal of Investigative Psychology and Offender Profiling under the title "The use of a linkage analysis as evidence in the conviction of the Newcastle serial murderer, South Africa." He returned to it in a 2012 chapter, "The Use of a Linkage Analysis as an Investigative Tool and Evidential Material in Serial Offenses," in the textbook Serial Offenders: Theory and Practice, where the Newcastle murders serve as the main worked example.
Later researchers cited the case as a point of reference. A 2019 study in Legal and Criminological Psychology by a team led by Jessica Woodhams, with Labuschagne among the authors, tested the principles of crime linkage on offense data from five countries, including case files of the South African Police Service. Its introduction lists State v. Sukude, Pietermaritzburg High Court case number 34/2006, among South African cases in which crime linkage was admitted as similar fact evidence.
In England, the researchers H. Fawcett and K. Clark used the case to ask how a jury might react to the same kind of evidence. Their study, published in the Journal of Investigative Psychology and Offender Profiling in 2015, recruited 22 people eligible for jury service in England and Wales and divided them into three mock juries. Each watched a 24-minute simulated trial, recorded in a replica courtroom and closely based on State v. Sukude, in which a defendant faced six charges, four of them murder. In the simulation the first two offenses were tied to the defendant by identification and other physical evidence, and the last two by linkage analysis alone, which a defense expert criticized. The judge's directions in the simulation copied those given in the real case.
The mock jurors were far more skeptical than the Pietermaritzburg court had been. They treated physical evidence, especially DNA, as close to conclusive, and regarded linkage analysis as circumstantial evidence that could not on its own show that a defendant was guilty. Some doubted that the shared features of the attacks were unusual enough to link them. They were wary of academic experts, and their views were often shaped by what they had seen in the media, including the idea of copycat crimes. Fawcett and Clark noted that linkage evidence had been accepted in South Africa in the Sukude case and in the 2011 case of State v. Mogale, but only by judges, since there is no jury in that system. They concluded that the method could become useful in courts in England and Wales, but that more research was needed first, and that it might be best presented to jurors as expert opinion rather than as proof of guilt.
Dates. The Sapa report gave the period of the crimes as February 2003 to January 2005. The judgment, The Mercury and Labuschagne's chapter all place the first attack in February 2004.
The arrest date appears as March 8, 2005, in most of the judgment and as March 9 in one passage.
Admissions. Labuschagne wrote that Sukude admitted the first two murders while being questioned by police. That admission was not evidence at trial, and in court Sukude denied all the charges.
No further offenses were attributed to Sukude in the published court record or the press reports of the trial.
Sukude was taken into custody in March 2005 and remained in custody through his trial. When he was sentenced in April 2006 he began five concurrent life terms, with the judge's direction that he serve at least 25 years before he could be considered for parole. His later whereabouts in the prison system have not been reported.
May 16, 1974: Themba Anton Sukude is born.
1990: the first of a series of criminal convictions the court later described.
February 2004: Veli Ncamo is killed in Newcastle; the woman with him survives.
October 27, 2004: Tebogo Mthethwa, a minibus taxi driver, is killed in Newcastle; the woman with him survives.
November 26, 2004: a man whose identity was not established is found dead in Newcastle.
January 8, 2005: Dhanpalan Govender is found critically injured in Newcastle; he dies about two days after the attack.
March 8, 2005: Sukude is arrested after a man he had spoken to reports their conversation to the police.
March 10, 2005: one survivor picks his photograph from a set of seven.
March 16, 2005: the other survivor identifies him at a parade.
2006: Labuschagne's article on the linkage analysis appears in the Journal of Investigative Psychology and Offender Profiling.
April 25, 2006: Judge Vivienne Niles-Duner convicts him in the Pietermaritzburg High Court on all nine counts and sentences him to five life terms and further terms, to run concurrently, with no parole consideration for at least 25 years.
April 26, 2006: The Mercury reports the judgment as the first use of linkage analysis to convict a serial offender in a South African court.
2012: Labuschagne's chapter on the case is published in Serial Offenders: Theory and Practice.
2015: a mock jury study in England based on the case is published.