A copper pot and a caustic soda ladle are held in the Criminological Museum in Rome. They belonged to Leonarda Cianciulli, a soothsayer in the town of Correggio in Emilia-Romagna, who used them on three women she killed in 1939 and 1940, and who is the reason the Italian press coined a name that has outlasted almost everything else about the case: the Soap-Maker of Correggio.
She was born at Montella in southern Italy in 1893, the child of a rape, and was told so by a mother who never forgave her for existing. She tried to kill herself twice as a girl. She eloped with Raffaele Pansardi against her mother's wishes, and her mother cursed the marriage and the children that came of it.
There were seventeen pregnancies. Three ended in miscarriage and ten of the children died young. Four lived. She went to fortune tellers looking for a way out of it, and was told by one that she would end her days in prison and an asylum, and by another that she would watch every one of her children die. She then set up as a soothsayer herself in Correggio, where she was well liked and much visited.
When Italy entered the war and her eldest and favourite son Giuseppe was called up, she decided that a life would have to be given for his.
Faustina Setti, an unmarried woman of her acquaintance, was told that a husband had been found for her in Pola, and was persuaded to write letters to her friends saying she had arrived safely and was well. She was given drugged wine and killed with an axe. Francesca Soavi was promised a job at a school in Piacenza and killed the same way in September 1940. Virginia Cacioppo, a former soprano, was promised work with a theatrical impresario and killed later that month.
What happened after the killings is why the case is remembered rather than filed. She dismembered the bodies, dissolved the remains in caustic soda in the copper pot, and made the result into cakes of soap, which she gave to her neighbours, and into a sweet tea cake, which she served to visitors and ate herself. Of Cacioppo she wrote that the flesh was the finest of the three and that the cakes were unusually good.
Cacioppo's sister-in-law did not believe the story she had been given about Florence, and went to the police. Cianciulli confessed at length and in detail. At her trial at Reggio Emilia in 1946 she was convicted and sentenced to thirty years in prison and three in a criminal asylum. She died of cerebral apoplexy at the women's criminal asylum at Pozzuoli on October 15, 1970. She had been told by a fortune teller that she would end her days in prison and an asylum, and she did.
There is a difficulty in the middle of all this, and it is worth naming rather than working around. Much of what is known about the killings comes from Cianciulli herself. She wrote a memoir in prison, An Embittered Soul's Confessions, and the details that make the case famous, the recipe, the texture, the judgement on Cacioppo's flesh, are her own account of her own crimes, written by a woman with every reason to be believed about the fact of the murders and no particular reason to be reliable about anything else.
This archive holds several cases that show what an account like that can be worth, and they do not all point the same way.
Henry Lee Lucas confessed to about six hundred murders after his arrest in Montague County, Texas in June 1983. A Texas Ranger task force was built around him, and police forces from across the United States came to Georgetown to put their unsolved cases to him. He cleared 213 of them. It was almost entirely false. In 1985 the Dallas Times Herald published payroll and cheque records placing him in other states on days he had confessed to killings, and the Texas Attorney General's own report the following year found he could not have committed the great majority. Investigators had shown him case files, driven him to the scenes and asked leading questions, and he had obliged. He was convicted of eleven murders and the credible figure is around three. The 213 cleared cases went back to the unsolved pile years cold, and some were never reinvestigated at all.
Ottis Toole began confessing in Florida at the same time, having travelled with Lucas for years after the two met at a Jacksonville soup kitchen in 1976. Between them they claimed hundreds of killings. Almost all of it collapsed. Toole recanted repeatedly, gave incompatible versions, and was demonstrably elsewhere for many of the deaths he described. He was convicted of two. The case the public attaches to his name, the killing of six-year-old Adam Walsh, taken from a shopping centre at Hollywood, Florida in July 1981, he confessed to several times and retracted several times, and he was never charged. The physical evidence that might have settled it, including a car and a machete, was lost or destroyed by the Hollywood police. In December 2008 the force closed the case and named him as the killer on the balance of what was left. He had been dead for twelve years.
Sture Bergwall, who took the name Thomas Quick, confessed to more than thirty murders while held at the Sater psychiatric hospital in Sweden under heavy medication and recovered-memory therapy. Between 1994 and 2001 he was convicted of eight, in trials where he pleaded guilty, where no forensic evidence connected him to any scene, and where no witness placed him there. In 2008 he told the journalist Hannes Rastam that he had invented all of it, for the attention and to keep the medication coming. All eight convictions were quashed between 2010 and 2013. The Swedish government's inquiry reported in 2015 and was critical of the police, the prosecutors, the courts and the treating clinicians alike. He was Sweden’s most notorious criminal for a decade and was described in the press as the country’s only serial killer. This archive keeps an entry for him because every reference work still lists him as one, and the entry exists to say that he is not.
Tommy Lynn Sells points the other way, which is why he belongs here. He was convicted in September 2000 of the murder of thirteen-year-old Kaylene Harris in Del Rio, Texas, a case that ended only because her ten-year-old friend Krystal Surles, whose throat he had cut, held her hand over the wound, crossed a field in the dark and wrote down what had happened because she could not speak. He then talked to investigators from a dozen states for years. Around twenty-two of the killings he described were corroborated. He gave figures for himself as high as seventy, and most of those have never been tested. One of his accounts undid a miscarriage of justice: in 1997 ten-year-old Joel Kirkpatrick was stabbed to death in Lawrenceville, Illinois, and his mother Julie Rea, who said an intruder had done it, was convicted of his murder in 2002. Sells described the killing in detail. Rea's conviction was overturned, she was acquitted at retrial in 2006, and she was later granted a certificate of innocence.
So a killer's own telling is not worth nothing and it is not worth taking on trust. What separates the four is not how convincing the teller was, and it is not whether a court believed them. Lucas was convicted of eleven murders and Bergwall of eight, and the confessions behind those convictions were worthless. Sells was corroborated on about twenty-two killings and never tested on the rest, and the difference in every instance was whether anything outside his account agreed with it.
Cianciulli's did. A relative went to the police because a story she had been told did not hold up. Three women could be named and were missing. There was a conviction at trial. There is a pot in a museum in Rome. What the memoir supplied was not the case against her but the detail everybody remembers, the recipe and the texture and the judgement on Cacioppo, and those are two different things. Lucas, Toole and Bergwall are what is left when the telling is all there is.