Evidence and guilt

| February 27, 2009

While DNA evidence may be sufficient to establish innocence, it is not sufficient by itself to establish guilt.

One of the worst misfortunes to befall a person is to be convicted of a serious crime of which he is innocent. Yet eyewitness statements may be unreliable and the prosecution may depend on forensic evidence to bolster its case, evidence whose reliability – with one exception, DNA traces – has never been subject to careful scientific investigation to see how reliable it is. Yet while DNA evidence may be sufficient to establish innocence, it is not sufficient by itself to establish guilt.

Nobody cares that the world's most famous detective, Sherlock Holmes, solved his cases through contrived coincidences, bending the truth and forensic clues that no modern police force would expect to help – you would hope. Nobody cares because Arthur Conan Doyle told some rattling good yarns. In fact when Holmes eventually died, readers insisted that he be brought back to life.

But what if the methods of forensic science used by today's police forces had little more validity than fictional yarns made up by Conan Doyle?

The US National Academy of Sciences has published a report which, it says in a news release, finds that:

[…] there is a dearth of peer-reviewed, published studies establishing the scientific bases and reliability of many forensic methods. Moreover, many forensic science labs are underfunded, understaffed, and have no effective oversight.

Forensic evidence is often offered in criminal prosecutions and civil litigation to support conclusions about individualization — in other words, to "match" a piece of evidence to a particular person, weapon, or other source. But with the exception of nuclear DNA analysis, the report says, no forensic method has been rigorously shown able to consistently, and with a high degree of certainty, demonstrate a connection between evidence and a specific individual or source. Non-DNA forensic disciplines have important roles, but many need substantial research to validate basic premises and techniques, assess limitations, and discern the sources and magnitude of error […]

As one example, it is widely assumed that an individual's fingerprints are unique but no study has ever been conducted to verify that. The reliability of expert evidence matching hair samples, bite marks, blood types and so forth to suspects has never been studied; nor has the rigour of matching bullets to the weapon that allegedly fired them.

As more becomes understood about the mechanisms of human memory, even eyewitness recall has been shown to be an uncertain guide to what really happened. Perhaps the most dramatic example is that of Australian memory researcher Dr Donald Thomson:

[He] was wrongly identified [in 1990] by a rape victim as the rapist. Thomson's alibi both exonerated him immediately and helped explain the false accusation: He was giving a live television interview at the time of the rape. Apparently, the victim had been watching the interview just before she was raped, and she confused the memory of his image with that of the rapist. Thus, source-monitoring failures can often be more harmful than retrieval failures: Fragments of real experience are accurately and vividly recalled but mistakenly attributed to the wrong person, location, or time, resulting in false memory.

Had Thomson not had the alibi (if for example the program had been prerecorded), he might have had great difficulty in resisting the charge, given the victim's vehement insistence that he was the rapist.

Testimony of expert witnesses may be equally problematical, as with British paediatrician Sir Roy Meadow whose testimony was instrumental in a number of women being convicted of murdering their baby children, before Meadow was found guilty of serious professional misconduct and struck off the medical register.

In the United States the Innocence Project, founded in 1992, has been able to exonerate more than 200 people, 17 of them on death row. Its success has depended largely on use of DNA evidence to override convictions based on less reliable forensic techniques.

DNA evidence is just about the only kind that has convincing statistical support. Even there though, beside the possibility of contamination at the crime scene, in transit or at the analysis laboratory, there are subtleties in interpretation that mean that it is never safe to convict on DNA evidence alone.

While wrongful conviction is a substantial issue in the United States, it has had a lower profile in Australia. The NSW Parliamentary Library published a briefing paper in 2006, "DNA Evidence, Wrongful Convictions and Wrongful Acquittals". It recognised that Australian justice systems are not infallible (and welcomed the intent of jurisdictions to link their DNA databases into a national system, an initiative that began in 1989 and is belatedly close to fruition).

The more we understand about forensic science the more complex it may become to determine that an accused person is guilty beyond all reasonable doubt.

If the National Academy of Science proposal is pursued so as to better understand how reliable the various types of evidence are, what does "beyond all reasonable doubt" then mean? That the odds are 20:1 that the accused did it? That they are 100:1? Or 1,000:1? Or do they have to be a million to one?

Considering all the complications involved in evidence it is perhaps surprising that the criminal justice system works as well as it does. But there is still every incentive to find improvements so that it works better.