Flawed science being used in Canadian courts
Bite-mark analysis has been shown to be bad science. So why is it still allowed in Canadian courts?
At a time when forensic science faces a reckoning, bite-mark analysis has come under intense scrutiny from scientists, lawyers and judges in the U.S., where at least 30 wrongful arrests or convictions have been based at least in part on this kind of evidence, according to New York-based Innocence Project.
The Star found bite-mark analysis is also used in Canadian courts. Here are five key points from our investigation:
á Judges appear to be rubber-stamping the use of such evidence by the prosecution or defence without questioning its scientific foundations.
á Published research over the past 10 years says the underlying assumptions of bite-mark analysis — that human teeth are unique and that skin can accurately record their impressions — haven’t been proven scientifically.
á One study showed that of 23 bite marks made on cadavers with the same set of teeth, no two were identical.
á In the case of the death of a 3-year-old Manitoba girl in 2006, three experts gave three different opinions on suspected bite marks found on her body.
á A former chief justice of the Ontario Superior Court says judges need to be better gatekeepers to prevent faulty science from entering the courtroom. Read the full investigation in Insight
“Tell me what happened, don’t leave anything out.”
The crime boss wanted Bradley Streiling to tell him everything about the death of 2-year-old Noah Cownden, Streiling’s stepson.
Streiling began, haltingly, to recount how Noah had slipped from the bathtub at his family’s Victoria home five years earlier.
Then, as a hidden microphone recorded, Streiling’s confession:
Filled with rage, he gave the boy a push. “I held him basically by the upper neck, lower jaw … and just hit him down a couple times.”
The boy’s “eyes glossed over, and he … kind of made a wheezing sound and never woke up again.”
The crime boss was actually an undercover Mountie who had just reached the culmination of a nearly five-monthlong sting. Police believed that Streiling, 29, had bashed Noah’s head against the floor, causing the little boy’s brain to bleed and swell so much that oxygen couldn’t reach it. They thought they had him. But the case against Streiling unraveled in an acquittal. The judge ruled Streiling’s confession had not been proven and she was persuaded by evidence that purported to show someone other than Streiling had dug their teeth into the boy’s shoulder, leaving a bite mark.
Practitioners of bite-mark analysis, known as forensic odontologists, claim that they can both identify bite marks on human skin, and link such marks to an individual’s teeth — often in cases of murder, child abuse or sexual assault. This kind of bite-mark analysis is subjective, relying on experts’ opinions.
Numerous studies and reports have found the underlying assumptions of bite-mark analysis — that human teeth are unique and that skin can accurately record their impressions — haven’t been proven scientifically.
Research shows that skin’s elastic properties make it a bad medium for accurately recording teeth impressions; no two bite marks made on cadavers from the same set of teeth are identical, one study found; and in another, experts couldn’t agree on whether an injury was made by teeth, suction or a bottle top.
One study found that of 89 bite marks made on cadavers with one set of teeth, none of the marks matched the set that created them when a computer program performed detailed measurements. The closest match was a different set of teeth in a randomized group of more than 400 other sets.
In 2016, scientists and engineers appointed by U.S. president Barack Obama arrived at this damning conclusion: scientific research “strongly suggests that examiners cannot consistently agree on whether an injury is a human bite mark and cannot identify the source of the bite mark with reasonable accuracy.” They also said the likelihood of bitemark analysis ever becoming scientifically valid was “low.”
This followed a major report in 2009 by the U.S. National Academy of Sciences that found no scientific studies support experts’ assessment that teeth marks on skin can identify a biter.
Study after study by scientists in the U.S., the U.K. and Australia culminated in a withering report published about two years ago in which 38 legal, medical and scientific scholars found the “rise and possible fall” of bite-mark evidence has highlighted the “weak scientific culture” of forensic science and the “law’s difficulty in evaluating and responding to unreliable and unscientific evidence.”
A Star investigation has found that as bite-mark evidence is being discredited in the U.S. — including its role in more than two dozen wrongful arrests or convictions — Canadian courts appear to be rubber-stamping its use.
The Supreme Court of Canada has stressed the importance of judges as gatekeepers for the courts. Just last month, during a Toronto sex assault trial, a judge, who said he had never before seen an expert in bite-mark identification, allowed one to testify without question.
Like hair microscopy and bullet lead analysis before it, bite-mark analysis is coming under intense scrutiny at a time when forensic science in general is facing a reckoning.
Patrick LeSage, former Chief Justice of the Ontario Superior Court of Justice, warned more than10 years ago that purportedly scientific evidence pointing to identity should not be used in criminal trials unless it has a “strong empirical and/or theoretical foundation.” In an interview with the Star, he said expert testimony, which is opinion evidence, must be carefully scrutinized by judges before it is admitted.
Judges can use hearings on an expert’s qualifications to also question the lawyers about their witness’s field of expertise, LeSage said.
His concerns about forensic science in general were echoed by Bruce MacFarlane, former Assistant Deputy Attorney General of Canada responsible for federal prosecutions, who wrote that warning flags should be raised when it comes to expert witnesses who rely mainly on their experience “rather than on an appropriate scientific underpinning.”
In the U.S., at least 30 wrongful arrests or convictions have been based at least in part on bite-mark evidence, according to the New York-based Innocence Project. The Texas Court of Criminal Appeals recently overturned the convic- tion of Steven Chaney, imprisoned for nearly 30 years for double murder. His conviction was based in large part on the testimony of two forensic odontologists. The appeals court said “the body of scientific knowledge underlying the field of bite-mark comparisons evolved in a way that discredits almost all the probabilistic bite-mark evidence at trial.”
“Bite-mark analysis is subjective speculation masquerading as scientific evidence,” said Chris Fabricant, a lawyer with the New York Innocence Project who represented Chaney. “It has no place in court.”
In Canada, three high-profile forensic odontologists (who are also dentists) have been frequently called on to give bite-mark evidence: Bob Wood, former head of dentistry at University Health Network in Toronto; Robert Dorion, responsible for the forensic dentistry department at the Laboratoire de sciences judiciaires et de médecine légale in Montreal; and Vancouver dentistry professor David Sweet. All have held leadership positions in the American Board of Forensic Odontology (ABFO), the field’s U.S.-based certification group, and all vigorously defend their work in bitemark analysis.
It was Sweet, director of UBC’s Bureau of Legal Dentistry (BOLD) forensic odontology lab, who was called in to assist with the autopsy of Noah Cownden after an odd-shaped injury was found on the boy’s shoulder.
Noah was precocious: at just 22 months old, he could spell his own name by saying the letters. The kid with blue eyes, a button nose and light brown hair loved hockey, which surprised his parents because no one in the family played. He was always in a hurry to keep up with his three older siblings.
On April 9, 2008, Noah suffered a massive head injury while alone in the care of his stepfather, Brad Streiling, who told investigators the little boy fell from the edge of the bathtub.
At the hospital, doctors found Noah had a subdural hematoma — blood pooling outside the brain. They cut away part of Noah’s skull to drain the blood and relieve the pressure. When that didn’t work, they removed part of the brain. But the swelling continued. Noah died a short time later, three days shy of his second birthday.
The following week, forensic investigators, including Sweet, stood over Noah’s body in the morgue at B.C. Children’s Hospital.
Sweet zeroed in on what looked to him like a bite mark on Noah’s left shoulder. He also found two other marks he thought could be bite marks — one on the boy’s left cheek and another on the back of his thigh.
Sweet measured the markings to see if the dimensions were similar to a human jaw.
“You don’t get a better case than this. Here’s the odontologist personally examining the marks, not just depending upon photos,” Sweet said in a recent interview. Sweet would not provide his reports but agreed to read parts over the phone.
He concluded the marks on Noah’s cheek and thigh were “suggestive of a human bite mark.” The injury on the boy’s shoulder showed more detail. Sweet saw several teeth “clearly outlined” and deemed the injury of “high forensic significance.”
He concluded the mark was a single lower arch of teeth on an angle pointing upwards, as if the biter had been standing behind the boy and leaned over the shoulder to make the bite. There were no marks corresponding to upper teeth, which Sweet said could have been because clothing had been in the way or the biter had no upper teeth.
“You see, you’ve got a dentist who understands development of the teeth and sizes and where they’re located so they can interpret the injury,” he told the Star.
Bite-mark analysis, used for more than four decades in hundreds of American cases, had a star turn in 1979 during Ted Bundy’s televised double-murder trial in Florida. Forensic odontologist Richard Souviron testified that bite marks on one victim’s body matched Bundy’s teeth, helping to secure a guilty verdict.
Forensic odontology is not recognized as a specialty by the American Dental Association or the Canadian Dental Association. Dentists seeking recognition of their credentials can apply to become diplomates of the American Board of Forensic Odontology (ABFO). There is no certification body in Canada.
Canadian courts have not tracked forensic evidence that is admitted, so it is unclear how such evidence has been used during trials or hearings, or how often. A recent study found that bitemark analysis was admitted in at least 14 Canadian cases involving serious crimes or child guardianship disputes. In each case, judges did not question the scientific validity of bite-mark analysis. Four cases took place after the damning 2009 National Academy of Sciences report.
“These findings suggest the current system isn’t working as it should,” said Jason Chin, a coauthor of the study and a researcher and lecturer at the University of Queensland in Australia.
“The problem seems to be that many practices have been accepted in the past, they seem superficially credible, and judges generally don’t seem to want to wade into the science.”
Bite-mark experts argue that recently strengthened standards and guidelines no longer allow definitive matching. Practitioners are now encouraged by the ABFO, the U.S.-based certification group, to come to one of three conclusions: the suspect can be “excluded” as the biter, “not excluded,” or the evidence is inconclusive.
While Canada’s top bite-mark experts defend their work, two acknowledged there are problems with the field’s scientific foundations: David Sweet, in a 2001 study he co-authored, wrote that there is a “lack of hard scientific evidence to support the assumptions made by forensic dentists when analyzing bite marks”; and Bob Wood told the Star that the scientific underpinnings of attributing a bite mark to a suspect are currently “weak.” He also said skin is a poor medium for accurately recording bite marks and is calling for a moratorium on attributing a bite mark to a specific individual.
Yet, Sweet and Wood challenge some of the recent U.S. research that found problems with bite-mark analysis. Sweet dismissed the President’s council study as “weak” and noted that U.S. prosecutors, the FBI and crime lab directors have criticized it, too. Wood criticized a study that showed bite-mark experts can’t consistently agree on which skin injuries are bite marks. He was one of the experts surveyed for the study and told the Star its design did not reflect how
bite marks are analyzed in real cases.
Sweet, who noted his role in helping to exonerate wrongly convicted American defendants, said he always attempts “to do the right thing and to be as conscientious and accurate as possible.”
Robert Dorion said it was wrong to say there is no scientific evidence showing human teeth are unique. He claimed there is research showing human skin can accurately record teeth and that he had done some of this research and had “demonstrated that at presentations.” When asked to provide it, Dorion sent the Star a list of bite-mark presentations he has given in the U.S. and France since 2012.
Police didn’t believe Noah Cownden’s massive head injury was the result of a fall. They suspected Streiling but needed proof. So four years after Noah died, investigators from several police forces, including the RCMP, resorted to a “Mr. Big” sting.
Mr. Big stings are controversial operations that entice suspects into joining criminal organizations, which in reality are creations of the police. Undercover officers posing as gang members befriend targets. The goal is to make the target comfortable enough to talk about the crime of which he is suspected.
The Supreme Court ruled Mr. Big confessions are admissible but only if the reliability of the confession and police conduct, among other things, are closely scrutinized by the judge.
Beginning in November 2012, officers conjured 52 scenarios designed to make Streiling talk. They took him on a boat to test-fire automatic weapons. They paid him to threaten a (fake) customs officer and drop bags of contraband (also fake).
In one scenario, a (fake) gang member pretended to beat up his pregnant partner while Streiling waited outside the room. The scenario was designed to show that the organization would not judge Streiling if he had committed similar “non-braggable” offences.
As the operation moved to its climax, police secretly recorded a meeting between Streiling and the head of the organization — “Mr. Big.” The boss told Streiling that investigators had more evidence implicating him in Noah’s death and offered a lifeline: a terminally ill member of the group would confess to the killing. Streiling just had to provide the details so that the fall guy could convince the police.
Streiling’s confession would play a large role at trial. So would evidence from Dr. David Sweet.
In 2014, a year before Streiling’s trial, and as criticism of bite-mark analysis mounted, the American Board of Forensic Odontology conducted a study in which nearly 40 forensic dentists with an average of 20 years experience were asked to analyze 100 photos. It found that the experts often could not agree if the photos were sufficient for analysis and often could not agree which injuries were bite marks.
“If two equally ‘qualified’ and experienced forensic odontologists look at the exact same injury and routinely come to opposite conclusions, the discipline is fundamentally unreliable, even at the threshold level of determining if the injury at issue is even a bite mark, to say nothing of attempting to associate or exclude any particular individual from having created the injury,” said the Innocence Project’s Fabricant.
Wide disagreement among bite-mark experts was on display in a prominent 2012 Manitoba court case involving the gruesome death of a child. Three experts gave three different opinions.
Jason Kines, 31, was on trial for the first-degree murder, aggravated sexual assault and sexual interference in the 2006 death of his girlfriend’s daughter, Venecia Audy.
The three-year-old had suffered a skull fracture, broken ribs, and a spinal injury. She had several suspected bite marks on her body, including one just above her vagina.
The murder case against Kines, who was not at home when Venecia died, hinged on bite-mark evidence from Sweet, a Crown witness.
During a hearing to determine the admissibility of this evidence, a defence lawyer questioned the science behind Sweet’s work, and Sweet acknowledged that there is no proof people’s teeth are unique. The judge allowed Sweet’s evidence into the trial.
Sweet testified that Kines’s “very unusual” teeth led to his conclusion that Kines was the “probable” biter, telling the court that he was “very confident” in his conclusion.
“Dr. Sweet’s opinion that Kines was the ‘probable biter’ was scientifically invalid because there was (and is) no evidence that bite-mark analysis works in the way he was claiming,” said researcher Jason Chin. “In other words, when someone says ‘probable biter’ in cases like Kines’s, how often are they right or wrong on average? Most importantly, what is the false positive rate — how often do they say ‘probable biter’ and the person was not the biter? Simply put, we do not know the answers to those questions.”
Under questioning, Sweet told the judge and jury that mistakes by bitemark experts had contributed to wrongful convictions in the U.S. that resulted in death sentences.
But the judge didn’t challenge the scientific validity of Sweet’s analysis. Instead, he seemed to rely on it to reach a decision. Because Sweet said Kines was the “probable” biter and not definitively the biter, the judge dismissed the jury and acquitted Kines. An appeals court ordered a new trial. Then, the case against Kines began to crumble. In a supplemental report, Sweet changed his analysis from Kines being the “probable” biter to saying he could not exclude Kines as the biter — a lower level of confidence. He told the Star he did this to address evolving standards for what bite mark experts like himself could say in court and because the Crown asked him to broaden his analysis to consider the general population.
The defence asked forensic odontologist Robert Dorion for his opinion and he disagreed with most of Sweet’s findings. Dorion concluded that all of the suspected bite marks had little or no value as evidence.
Dorion said that one of the marks Sweet had attributed to Kines could not have come from the accused “without dislocating the jaw.” He also said three injuries suggested Venecia had bitten herself.
A third bite-mark analyst, hired by the Crown to review Sweet’s work, agreed with Sweet’s assessment of the majority of the injuries, but came to different conclusions on four others.
A final piece of evidence seemed to undermine the bite-mark evidence: DNA recovered from the bite marks on Venecia’s body revealed two profiles of people other than Kines, according to a Crown attorney with the Manitoba Prosecution Service.
The Crown withdrew the charges and Kines walked. (Kines was found guilty of failing to provide the necessities of life and was sentenced to one year, less time served).
Whoever killed Venecia Audy has not been brought to justice.
It’s not clear if what happened in the Kines case and its implications for bitemark evidence was noticed by any other court.
Former Chief Justice of the Ontario Superior Court Patrick LeSage says judges must be better gatekeepers to ensure faulty science doesn’t make its way into the courts.
“Judges have to be constantly aware. Judges don’t have to be scientists, but they need to know there are pitfalls and areas they can easily get lulled into,” said LeSage, who in 2005, presided over an inquiry into the wrongful conviction of James Driskell, a Winnipeg man sentenced to life imprisonment for murder, a conviction that hinged partly on nowdiscredited hair microscopy evidence.
In May 2013, after Brad Streiling was arrested, David Sweet examined five pairs of anonymous dental casts.
He used a computer to help him compare the edges of the lower teeth of each cast to a photograph of the injury on Noah Cownden’s shoulder. He concluded he could exclude all, except dental cast 3. This was verified, he said, by two other forensic odontologists in B.C.
Streiling’s teeth were among those Sweet excluded.
In an interview with the Star, Sweet said a decision to exclude a suspect can be made with “100 per cent certainty,” a claim the Innocence Project’s Fabricant calls “absurd.”
“Nothing in science — not DNA, not nuclear physics — is ‘100 per cent certain.’ Nothing. That a dentist would make such a claim, despite never even having such an ability tested, is all the more absurd,” Fabricant said.
There was no hearing on the admissibility of bite-mark evidence in Streiling’s trial. Neither the defence nor prosecution, which called the evidence, raised any concerns, and the judge allowed it in.
Streiling, who could not be reached for comment, testified that his confession to Mr. Big was false. He believed that if he didn’t confess, he would be kicked out of the organization.
Streiling’s lawyer Martin Allen told the Star that the defence theory was that Noah had a prior brain injury that made him “susceptible to a traumatic and fatal bleed from a relatively minor blow.”
In her ruling, the judge said the Crown’s medical evidence wasn’t strong enough to overcome weaknesses in its case.
The judge ruled Streiling’s confession was “vague and lacking in detail” and that it was “too unreliable to accept as a true admission of guilt.”
“Police must be aware that when they undertake Mr. Big investigations, they are more likely to be successful in obtaining a conviction on the strength of such confessions in cases where there is independent confirmatory evidence…and not evidence that is subject to differing opinions and interpretation such as expert medical evidence,” she noted in her judgment.
But Sweet’s evidence gave the judge comfort that Streiling was the wrong man. She was convinced the mark was a bite and that it came from someone else.
“Since the evidence does establish that Noah was bitten by an adult other than Mr. Streiling, I must consider that fact as an important part of my analysis,” she wrote in her judgment. “Absent any other explanation, the bite evidence suggests that Noah suffered abuse at the hands of a person other than Mr. Streiling.”