On November 2, 2007, a brutal murder shocked Italy. Meredith Kercher, a 21-year-old British student, was found dead with signs of rape in her bedroom in the house in Perugia where she was attending an Erasmus year. She had received 46 stab wounds. Suspicions soon pointed to her flatmate Amanda Knox and to Knox’s Italian boyfriend, Raffaele Sollecito. The discovery of traces of Knox’s DNA on a knife where Kercher’s blood was also found and the detection of some of Sollecito’s DNA in the victim’s bra were also decisive in seeing the couple sentenced to 26 and 25 years in prison. Case closed. Forensic science once again played a crucial role in a judicial decision.
However, in 2011 the couple was acquitted. Experts Stefano Conti and Carla Vecchiotti discredited the main tests because the investigation of the Italian scientific police “had not followed the international protocols for the collection and processing of samples.” The amount of DNA of the victim found on the edge of the alleged crime knife “was too small to reach definitive conclusions” and the DNA from Sollecito in the bra, in which traces of DNA from others were also detected, was also inconclusive. Conti and Vecchiotti pointed to the possible contamination of the evidence. This resolution, which had an impact on the international and scientific community, revealed an oft-ignored reality: forensic science also makes mistakes.
The errors in the Knox case went round the world, but this is not the only example. “Unfortunately, this is too frequent. It is not just those cases that could lead to imprisonment that matter. The fact that an innocent person is investigated and taken to a police station because of forensic malpractice is already doing things wrong,” says Fernando Verdú, a forensic doctor, Professor of Legal Medicine at the University of Valencia and former president of the Ibero-American Society of Medical Law at OpenMind.
Reducing the weight of forensic evidence
“We forensic scientists have to recognize our limitations, which are many. We must reduce forensic medicine to the level it really should have and also reduce the weight that the judicial system places on expert evidence, which, unfortunately, is excessive,” says Verdú, who is also director of the Master’s Degree in Forensic Medicine at the University of Valencia. A survey carried out for New Scientist magazine of specialists in DNA analysis revealed these two results: 10 out of 12 of these analysts strongly believed that the police had a high level of faith in DNA findings and did not understand their limitations, and 9 out of 13 believed that the same occurred in judicial decisions.
Enrique Villanueva, Professor of Legal Medicine at the University of Granada and president of the National Commission of Forensic and Legal Medicine also acknowledges the great weight given to the opinions of forensic experts in legal trials: “It ends up being practically impossible to make another opinion prevail.” This forensic doctor also explains the cause: “The confidence that the judges have placed in forensic doctors derives from the deserved prestige created over a century of good work. But the title does not ensure infallibility. The judgment of the forensic doctor is to the evidence what notarization is to a will,” Villanueva explains to OpenMind. That is, it provides a seal of approval.
No reliability in visual comparisons
In spite of its limitations and the legal errors committed by those who collect the samples of contaminated or misinterpreted DNA, this technique represented a turning point in the reliability of forensic science. For decades, the forensic scientists’ procedure was based on the visual comparison under the microscope of hair, fibers, bullets, tools, fingerprints or bite marks, looking for the physical similarity with the evidence collected at the scene of the crime. This resulted in a clear inaccuracy since there has never existed—and still does not exist today—a global database in which to compare this type of evidence.
“Studies have shown how experts cannot even say whether a bite mark is human, let alone recognize who left it,” said Alicia Carriquiry, head of the Forensic Science Center of Excellence at Iowa State University. Although in other areas such as footprints or the analysis of tool marks, pairings are a bit more precise, they remain unclear.
The US organization Innocence Project was created precisely for that purpose: to locate and rectify with DNA evidence those errors in forensic evidence that led to the imprisonment of innocent people. In two decades, they have been able to exonerate 349 people, of whom 20 were sentenced to death, and have found the true perpetrator of the crime in 149 cases, according to the information offered by this organization on its website. “It is true that there are many people convicted on the basis of expert evidence that turned out to be false. Very few tests lead us to the absolute truth. We doubt the experts less than we should. DNA has come to help a lot, but it does not solve everything,” says Villanueva.
The solution to do it the other way around
In addition to the lack of accuracy of this type of visual comparison ‘by eye’—now fallen into disuse or with the support of the DNA behind it—some experts seek to focus on those techniques that are widespread, but not as accurate as one would like, such as the distance of the gunshots, the evolution of the hematomas or the time that a body has remained in the water. One of the most debated is the date or time of death. “It is impossible to give it exactly. Each cadaver evolves in its own way, depends on many physical and chemical processes, from a myriad of factors,” describes Verdú. The expert at the University of Granada also recommends caution: “I can assure you that a yellow bruise is not recent, but I will be very reckless if I say that it is 10 days old. The date of death will always be approximate, always with a wide margin of error.”
One of the solutions proposed by the forensic doctor from the University of Valencia is to do the procedure in reverse. “If I indicate that the time since death is between 12 and 24 hours, the investigators are going to focus on the suspects who were with the victim during that period of time,” he explains. “However, the logical way to do it would be the opposite: the police start the investigations with a wider range of possibilities and when they have a suspect that fits in a time slot, they can go to the forensic pathologist to determine if it is possible or not,” he concludes. Professor Itiel Dror, a neuroscientist at the University College London, said that things could be vastly improved if investigations, instead of being led by the forensic samples, were instead focused on the suspects in the case.
Comments on this publication