Forensic science is largely a faith-based discipline whose experts testify first and worry about verification later, according to a leading U.S. law and social science scholar.
“There’s a lack of data but an enormous amount of faith in it,” Michael Saks told a Toronto conference on Nov. 23 to mark the 20th anniversary of Canada’s Association in Defence of the Wrongly Convicted. The association has played a significant role in the exonerations of 18 wrongly convicted Canadians.
Forensic science is the second-largest cause of erroneous convictions after misidentifications, said the Arizona State University professor.
At least three forensic fields “are now deceased” with their removal from the “Pantheon of science” in the United States, according to Saks: voice print identification, comparative bullet lead analysis, and almost two dozen arson indicators.
Bite-mark identification and microscopic hair comparison will likely suffer the same fate, Saks told more than 150 people, including victims of wrongful convictions, during a panel discussion.
“But the point is that none of these were kicked out of courts by judges,” he said. “These were determined externally to lack a scientific foundation and removed after decades of coming into court” and being “cheerfully admitted.”
Change will come, Saks said, but only through a number of policy choices:
• Adopting quality control measures to prevent “motivated perception” in which forensic examiners interpret the evidence to support the expectations of investigators and prosecutors.
• Sending random samples of crime laboratory examinations to independent laboratories for verification.
• Determining laboratory error rates and requiring them to be entered as evidence.
• Ensuring the use of weak-pattern comparison types of forensic evidence is for exclusion only and not for identification unless there’s sufficient validation research.
Fellow panellist Barry Scheck, a law professor at the Benjamin N. Cardozo School of Law in New York City and cofounder of the Innocence Project, said voice-print identification was discredited after the United States National Research Council convened scientists who examined the literature “and basically said, ‘You can’t do this.’”
The prestigious research council also found serious flaws in composite bullet lead analysis.
There’s now a “duty to correct” false presentations in the forensic scientists’ code of ethics, according to Scheck.
As for arson, infamous cases like the 2004 execution in Texas of Cameron Todd Willingham in the burning death of his three children brought fire investigation techniques into question, he said. The National Fire Protection Association has corrected many “phoney visual cues that so-called arson examiners have been using,” said Scheck.
And bite-mark forensics will never be validated, according to Scheck.
A whole set of forensic disciplines may eventually be scientifically verified once again, “but it’s going to look nothing like the cases that are already in the books where people have been convicted,” said the defence attorney.
Saks cited the work of psychologist Itiel Dror. He gave five fingerprint examiners pairs of prints they had identified as matches in previous cases. But researchers told them they were actually from the infamous FBI misidentification of Brandon Mayfield, a man wrongly linked to the 2004 Spanish train bombings. Three experts said the fingerprints weren’t matches. One said they were inconclusive. Only one identified the prints as from the same person.
Forensic examiners protest they can’t look at samples without knowing something about the case, said Saks. That’s fine, he said, but they should first be examining the samples blindly. They can then be gradually unmasked with additional information.
Scheck said even fingerprint experts acknowledge a false-positive rate of 0.1 per cent.
Nonetheless, fingerprint identification databases “can be a gold mine” for generating leads on alternate suspects when lawyers are trying to clear the wrongly convicted, said Scheck.
Saks advised defence lawyers to learn the weaknesses of specific techniques. “I have heard experts testify to things that their own field, in its textbooks, says they cannot do.”
The problem is lawyers can’t be experts on everyone’s expertise, Saks added. He advised lawyers to band together in knowledge networks.
He recommends interviewing the examiners. “Ask them to tell you everything they can tell you about how they reached the decision that they reached.”
If at all possible, lawyers should have a consulting expert of their own, he added.
Once in court, Saks advises lawyers to seek to exclude pattern comparison evidence like bite and tire marks and handwriting; ask the judge to order a blind examination of the evidence, which often involves getting into the case early before the samples have been submitted to the expert; seek an order placing limitations on what the expert can say about the inferences; request a ban on certain overpowering or exaggerated conclusions such as declaring something to be a match; and try to limit the expert’s testimony to things for which there’s an empirical and tested scientific basis.
Panel moderator and prominent Toronto criminal lawyer Alan Gold said defence counsel share some of the blame for not studying the increasingly robust scientific literature relevant to their cases. “There is now no excuse whatsoever,” said Gold.
In an earlier conference speech, Stephen Bindman, special adviser on wrongful convictions at the federal Justice Department, gave what he called a simple message: Canadian Crown prosecutors care about wrongful convictions, too.
“I think increasingly the Crown community like everyone else recognizes that wrongful convictions are a part of our system. They are not just isolated one-offs,” said Bindman, who stressed he was speaking for himself and not the government.
Nevertheless, Bindman added that “not everyone gets it and, yes, there will be legal fights that many of us wish wouldn’t have to be fought.”