As Amanda Knox walks free, now DNA evidence is on trial

by jeeg 4. October 2011 22:41

The sensational acquittal of Amanda Knox and her co-defendant Raffaele Sollecito should serve as a warning to prosecutors the world over who rely on hi-tech DNA evidence to make their case: be careful what you wish for. 

Because it is clear that DNA, the very stuff of life, does not always provide the ironclad evidence that modern forensic scientists insist it does.

With no obvious motive, no independent witnesses and no confessions, the Italian prosecutors had to rely on genetic evidence found at the crime scene to convict Ms Knox  and Mr Sollecito.

The original conviction, now overturned, relied on traces of Mr Sollecito’s DNA being found on the victim Meredith Kercher’s brassiere clasp, together with traces of DNA from both defendants being found on the knife allegedly used to slash Kercher’s throat.

Although there has been no explicit statement, it is clear that the appeal jury believed the defence case that this supposedly watertight DNA evidence was dubious at best, and downright fabricated at worst.

These are early days (and this case will be picked over for years), and one trial does not prove anything about the worth of any one forensic technique, but the Kercher case shows yet again that the ‘genetic fingerprint’ is as open to interpretation, bias and error as any previous means of placing a suspect at the scene of the crime and moreover proving that he or she did it.

In the case of the Kercher murder, the defence argued, presumably successfully, that there was no proof that Ms Knox and Mr Sollecito’s DNA was on the bra strap and knife, as the prosecution alleged.

This seems odd; surely DNA is either there or it is not? And isn’t every sample of human DNA unique?

Well, yes and no. Since the British geneticist Alec Jeffries pioneered the technique of forensic genetic fingerprinting in the 1980s, there is no doubt that hundreds of guilty murderers, rapists and thugs have been prosecuted successfully on DNA evidence (and a lesser number acquitted using the same technique). But it is not entirely failsafe.

There is, for a start, the issue of contamination, a factor it seems in the Kercher case. No one argues that the victim and the accused knew each other, were in the same places at the same time.

Any object that was in their vicinity could easily contain traces of the DNA of everyone involved.

The police are supposed to guard against this, but, again, no system is entirely foolproof.

Knives and fabric samples are put in cellophane bags and sometimes someone will make a mistake, allowing two objects to touch each other for instance.

In this case it seems that the case made by the defence, that there were errors made by the police and investigating authorities regarding the analysis of forensic material was enough to sway the appeal court.

But in fact the potential flaws with DNA evidence run a lot deeper than that.

We often hear that ‘DNA evidence’ shows that the chances of a particular person being innocent are ‘less than one in a million’ or some even more outlandish figure. 

Typically, a DNA sample found on a murder weapon will be said to match the DNA of the suspect to the extent that only one person in one million would have the same profile. Case closed?

Not really. This is the classic example of the ‘prosecutor’s fallacy’. If a particular genetic profile is held by one person in, say, one million this means that in a country the size of the UK (which has a population of 60m) 60 people will provide a perfect mach for the  DNA on the knife.

Thus, all things being equal, the chances that the man in the dock being innocent are not one-in-a-million against, but 59-in-60 for; rather different.

It gets worse. DNA evidence puts a person 'at the scene of the crime'; it does not prove they committed the crime.

Several people have been wrongly convicted of murder or robbery in the US after their DNA was found at the scene of, say, a heist or grocery store hold-up when in fact they had merely been there along with hundreds of other people.

DNA evidence can be used to draw a false inference of a link between known criminal background, a particular crime and a suspect being in a particular location.

DNA evidence is not very time-sensitive: in the case of  Madelaine McCanns tentative (and as it turned out spurious) DNA evidence 'placed' the missing girl’s  body in the hire car used by her parents after she disappeared.

But of course this could have come, retrospectively, from her siblings, her clothes, her toys or her parents.

Worse still DNA evidence is highly subjective.  In the US state of Georgia a man called Kerry Robinson was convicted a few years ago of gang rape: in an excellent New Scientist investigation last year DNA evidence from the crime scene plus Robinson's DNA profile was shown to 17 'blind' analysts with no contextual information: the 17 experts were hugely divided - 12 said the suspect could be excluded.

Close to home, Sean Hoey was cleared, in 2007, of the 1998 Omagh bombing.

This was the first time defence lawyers in the UK had successfully challenged 'infallible' DNA evidence in court, in this case spuriously showing that Hoey had touched the timers used to detonate the bomb.

In this case, tiny amounts of DNA were 'amplified' to generate a dubious 'profile' which 'matched' that of the defendant.

‘DNA’ has taken on a spuriously totemic status in many courts, a sort of divine sword of truth against which no one mortal dare argue.

No one can doubt that genetic evidence is a hugely important and valuable addition to the law’s arsenal.

But it is not infallible, and is subject to the same biases and random human errors as any other form of evidence.

Michael Hanlon, Daily Mail 



Comments are closed
Log in