Lead Articles

Can we really rely on Forensic Science?

Forensic science has come a long way since the days of Stephen Lawrence, a young man the mystery of whose savage, racist murder in 1993 only reached a conclusion this week. Forensic science now enjoys much higher standards of evidence handling and more sensitive technology, and these advances have evidently played major roles in bringing about the convictions of several high-profile criminals. Still, as with all things scientific, the reliability of forensic technology has always been in question.

In the case of Stephen Lawrence, Gary Dobson and David Norris were sent to jail on the basis of evidence which, the defence argued, could have been the result of contaminations during evidence storing. The findings that damned Dobson and Norris involved fibres from Lawrence’s clothes on Dobson’s jacket and Norris’ sweater, but perhaps more importantly, hair on Norris’ corduroy trousers and a 0.25mm by 0.5mm bloodstain on the collar of Dobson’s jacket — all of which were genetically matched to the deceased. Three tiny old flakes of Lawrence’s blood were also found in the evidence bag that held Dobson’s jacket, and the defence saw this as an indication that the bloodstain on Dobson’s clothes could have come about due to contaminant old flakes of blood dissolving during a saliva test.

The forensic team successfully ruled out this possibility when they showed that exposing said old flakes of blood to the chemicals used in a saliva test did not lead to them staining or absorbing into the fabric. The scientists were of the position that the stain was caused by wet blood that had been shed during the attack. Furthermore, the techniques used in this re-examination had already crucially contributed to the conclusion of other murder investigations such as those of Vikki Thompson, Damilola Taylor and Rachel Nickell. In the end, the evidence for the conviction of Dobson and Norris looked solid, and the jury’s decision reflected thus.

Yet, the field of forensic science has been plagued by much more than just courtroom debates. Aside from the risk of contamination that comes with poor evidence handling, the overall dubiousness of some of its practices has also been hauled into the public limelight. A 2009 report by the US National Academy of Science (NAS) suggests that the field needs an independent body to uphold higher standards of practice and traditional methods such as fingerprint and hair analyses cannot match evidence to individuals with certainty since not enough peer-viewed studies have established their effectiveness.

Downright horror stories such as those of wrongly-convicted Steven Barnes illustrate just how dangerous it can be to put too much trust into forensic science. In spite of shaky eye witness accounts and little other evidence, Barnes was falsely convicted for the rape and murder of a high school student in 1985, mainly because, according to a forensic expert, the soil on his truck tyres resembled the soil on the crime scene and an imprint similar to the fabric pattern on the  jeans the victim had been wearing at the time of death could be found on the outside of that truck. It was only after 20 years in prison that DNA tests proved that Barnes was innocent, leading to his exoneration.

Cases of this ilk go to show that sometimes forensic evidence is perceived to be conclusive, in spite of a dearth of other evidence available. In fact, with shows like CSI painting a rather more crass and idealised picture of an otherwise fairly cautious and reserved scientific field, jurors might find themselves willing to blindly believe any forensic evidence that they are presented with.

The aforementioned NAS report calls DNA analysis, with its low false positive and false negative rates, the exception to the rule. But apparently even DNA analysis cannot be fully trusted, as is the moral of last year’s conclusion to the Amanda Knox saga.

Knox and her boyfriend, Raffaele Sollecito, were being detained and trialled for the murder of her flatmate Meredith Kercher; during the course of this much-publicised Italian courtroom drama, the police brought in forensic evidence in the form of Sollecito’s DNA on Kercher’s brassier clasp and DNA from both defendant’s on the knife that was apparently used to kill the deceased. In spite of this, the defence seemingly managed to argue that said evidence could have been caused by non-direct means; indeed, because Knox was Kercher’s housemate, and Sollecito, Knox’s boyfriend, DNA was probably constantly being exchanged amongst the three and Sollecito’s DNA could simply have been passed on to Kercher’s brassier without him ever having touched it. The judge deemed the DNA evidence “not reliable” and Knox and Sollecito were duly acquitted.

Another example of an acquittal in spite of DNA evidence is that of Sean Hoey. Hoey was trialled for the 1998 Omagh bombing in 2006 because his DNA had been found on items such as the timer that was apparently used to trigger the bomb. The prosecution was build around this evidence and that was perhaps its downfall, due in part to the inherent unreliability of Low Copy Number, or LCN, the method used to detect the evidence. This technique works by amplifying extremely small DNA samples to make them more detectable, but with the very possible risk of amplifying contaminant DNA in the process. Though the technique has led to some convictions in previous cases, it has not gained significant accreditation from the wider scientific community. Hoey’s defence banked on these doubts and successfully contested the DNA evidence, which eventually saw the defendant cleared of all charges.

Forensic science is not infallible, and neither is DNA profiling, which is why we should always look beyond CSI’s chimera, and question the evidence to its tiniest detail. But the proponents of forensic science are not unaware of its problems, and just like every other scientist of every other discipline, forensic scientists will continually seek to perfect it. Forensic science is flawed, but it has nonetheless revolutionised the face of criminal justice in many countries, and mostly for the better. It is hopefully only a matter of time before scientists overcome the challenges outlined in this article.

Eric John

Lead ArticlesScience
One Comment
  • The dangers of forensic evidence | Ones to Watch
    10 January 2012 at 19:58
    Leave a Reply

    […] article here. Share this:FacebookTwitterShareRedditDiggStumbleUponPrintEmail Tagged as: Davis Norris, Eric […]

  • Leave a Reply