In November 1983, a fifteen-year-old named Lynda Mann was raped and strangled on a footpath through the English village of Narborough, Leicestershire. Nearly three years later, in the summer of 1986, a second schoolgirl, Dawn Ashworth, was killed in the neighboring village of Enderby. Investigators were certain one man had done both. When they questioned Richard Buckland, a local seventeen-year-old who had let slip details about Ashworth's body that had not been made public, he confessed to her murder. He denied, just as firmly, ever touching Lynda Mann.
Police wanted one confession to cover two killings, so they turned to Alec Jeffreys, a geneticist at the University of Leicester who had discovered, two years earlier, how to read the repeating stretches of DNA that make every person's genome identifiably their own. The year before the Narborough case, he had used the technique for the first time in a legal dispute, to settle a family's contested immigration status; nobody had tried it on a criminal suspect. Jeffreys compared semen recovered from both bodies against Buckland's blood. The two crime samples matched each other. Neither matched him. Because the method had never been used in a courtroom, Britain's Forensic Science Service ran its own tests before anyone acted on Jeffreys' result, and reached the same answer. Buckland was released. He is remembered as the first person a DNA test is recorded as having cleared of a murder charge.
With one confession gone and the killer still free, Leicestershire Constabulary attempted something no police force had tried before: it asked every adult man in three villages, about five thousand of them, to volunteer blood or saliva. A blood-group test narrowed the pool to the roughly one in ten men whose type matched the semen from both killings, and DNA profiling checked each of those men in turn. A local bakery worker, Colin Pitchfork, arranged for a friend named Kelly to take the test under his name: he put Kelly's photograph into his own passport and taught him the details of his domestic life so the impersonation would hold. Kelly gave blood on Pitchfork's behalf on January 27, 1987. It worked until Pitchfork, growing careless over drinks at a pub, let the substitution slip to someone who told the police.
Pitchfork was arrested, and DNA matched him to both murders and to two earlier sexual assaults on teenage girls nearby. He pleaded guilty at the Crown Court in Leicester on January 22, 1988, and was sentenced to life imprisonment. Passing sentence, Mr Justice Otton told him what the new test had changed.
A confession has always been the law's favorite kind of evidence, since it comes straight from the person being asked. It is also, courts have relearned again and again, one of the easiest kinds of truth to manufacture under pressure: through exhaustion, suggestion, or a frightened teenager's wish to go home. The test that cleared Buckland didn't know what anyone had said in an interview room. It read a molecule that had been sitting in an evidence sample since before the technique existed to read it, and it gave the same answer no matter who was asking.
A confession says what a person is willing to tell you. In 1986, for the first time, the blood itself had a way to answer back.