After serving 28 years in prison, Steven Chaney walked away a free man last Monday when a Dallas judge overturned his 1987 murder conviction. The clincher that sent Chaney to prison nearly three decades ago? Bite mark testimony, given at his trial by two forensic odontologists. The clincher that secured his freedom? Discredited bite mark testimony.
The outcome of Chaney’s case is yet another notable strike against the controversial practice of using bite marks to secure convictions. For decades, testimony from forensic dentists—who inspect the injuries of victims and attempt to match them to the dental patterns of alleged perpetrators—has been admissible in court. Often, this testimony is the prosecutor’s only physical evidence.
Chaney was on trial for the murder of John Sweek, a drug dealer found stabbed to death on his kitchen floor. Chaney, a construction worker, had been one of Sweek’s customers. During the trial, Homer Campbell, a forensic dentist from Albuquerque, told the court that there was a “reasonable dental certainty” that the bite marks on Sweek’s arm came from Chaney. Jim Hales, chief dental consultant for the Dallas County Medical Examiner’s Office, piled on with an alarming statistic: that there was a “one to a million chance” someone other than Chaney was the biter. Even though Chaney had nine alibi witnesses, the jury placed considerable stock in the word of these experts. One juror, when asked why he voted for Chaney’s guilt, said, “The bite mark.”
But now, in an affidavit filed with the court, Hales has admitted what critics of bite mark evidence have been saying for years: even an expert can’t reliably match bite marks to teeth.
Conclusions that a particular individual is the biter and their dentition is a match when you are dealing with an open population are now understood to be scientifically unsound. Under today’s scientific standards, I would not, and could not, testify to a reasonable medical/dental certainty as I testified at the time of trial nor could I testify that there was a ‘one to a million’ chance that anyone other than Mr. Chaney was the source off the bite mark.
Chaney’s lawyers—Julie Lesser, with the Dallas County Public Defender’s Office, and Barry Scheck of the Innocence Project—challenged the conviction, citing Texas’s junk science law, passed in 2013, which says that a conviction can be attacked in a writ of habeas corpus if there is new science that contradicts the science that was used at trial.
The truth is, there was never any conclusive data or rigorous studies to back up bite mark evidence, which has been under fire from scientists and defense lawyers ever since it was first allowed in court in 1974. Tests of bite mark evidence reliability have found error rates between 12 and 64 percent, but since no court ever ruled against its admissibility, it was allowed.
Bite marks are nothing like DNA profiles.
Forensic odontologists sometimes can’t even agree on whether marks found on skin—a malleable, inconstant medium—came from teeth or not. The first official red flag about bite mark evidence came in 2009 when a report from the National Academy of Sciences said, “The scientific basis is insufficient to conclude that bite mark comparisons can result in a conclusive match.” Around that time a husband and wife team of researchers at SUNY Buffalo began doing research on bite marks using cadavers, and after more than a dozen studies, they found that “statements of dental uniqueness with respect to bitemark analysis in an open population are unsupportable.” In other words, bite marks are nothing like DNA profiles—and there are certainly no statistics to back up accurate comparisons between sets of teeth, like Hales did when he said there was a “one to a million chance” anyone but Chaney was the biter.
It was significant that Dallas County DA Susan Hawk concluded that “the bite mark evidence that was critical to [Chaney’s] conviction has been discredited”—he is now the twenty-sixth person to have been wrongly convicted or indicted based on bite mark testimony—but county prosecutors knew about his case for months. All summer long the Texas Forensic Science Commission—which for the past five years has been blazing a trail of state-wide criminal justice reforms via numerous investigations of labs and forensic disciplines—has been looking into bite marks after a complaint was filed by Chaney’s lawyers, who asked the commission to “exercise its statutory mandate to investigate and report on ‘the integrity and reliability’ of bite mark evidence.” The FSC has already held one meeting to look into bite marks—last month in Dallas—and Chaney’s name came up often. His name will come up again when the FSC convenes again next month in Fort Worth.
One of the things Chaney’s lawyers asked the FSC to do is go back and vet cases where bite mark testimony was used in Texas in the same way the FSC has been re-investigating old hair microscopy cases. When it does so, the commission will find other troublesome Texas cases, including three that Campbell (now deceased) and/or Hales handled at the same time as Chaney’s.
One of those cases involved two men convicted for the rape and murder of Juanita White in Waco in 1986. When investigators found what they believed to be bite marks on White’s body, they took a dental mold of a suspect named Calvin Washington and drove it to Dallas for Hales to inspect. His conclusion? Washington’s teeth matched the wounds on White’s body. But the story didn’t end there. Investigators began to suspect Washington had an accomplice, a man named Joe Sidney Williams, and they made a mold of his teeth too. This mold—along with Washington’s mold and White’s autopsy photos—was sent to Campbell, who saw things differently than Hales: Williams was the biter, not Washington. Prosecutors chose to go with Campbell’s identification, not Hales’s, and in August 1987 Williams went on trial. The only physical evidence were the bite marks. Campbell identified four of them on White’s body and said Williams’s teeth were consistent with an injury on her hip. “The research states that there are no two people that have the same position [of their teeth],” Campbell testified, though no such research has ever been done. Williams was found guilty, as was Washington in a later trial where almost the same evidence was presented.
But both Campbell and Hales were wrong, a fact not found out until 2000, when the semen in the rape kit was compared to the DNA profile of another man. It matched the new suspect and Washington was freed. (Williams had been freed in June 1993 because testimony from a jailhouse informant had been ruled inadmissible.) The two men served a total of 19 years in prison for a murder they had nothing to do with—all based on bogus bite mark testimony.
The third case is even more troubling because it involved an execution. The defendant’s name was David Spence, and he was, oddly enough, Juanita White’s son. (For more on this labyrinthian case read “The Murders at the Lake.”) Spence was convicted in two trials, in 1984 and 1985, of the murders of three Waco teens and given the death penalty. The only physical evidence against him: bite marks on the bodies of two of the victims. The expert who testified: Homer Campbell. Spence, Campbell said, was “the only individual” to a “reasonable medical and dental certainty” who could have bitten the women. According to jurors, Campbell’s words were powerful. “We had life-size pictures of the marks and a cast of [Spence’s] teeth brought into the jury room,” remembered one juror afterward. “The testimony—‘everyone’s bite mark is different, like a fingerprint’—was very convincing.”
Spence’s appellate lawyers tried to attack Campbell’s methods with other forensic odontologists. One, Thomas Krauss, a former president of the American Board of Forensic Odontology (ABFO), said Campbell’s methodology was “well outside the mainstream.” Krauss helped the lawyers set up a blind panel of five odontologists to analyze the autopsy photos and vet Campbell’s work by comparing the marks with dental molds from Spence and four other subjects. The results were astonishing. Though the five experts identified several patterns that were possibly bite marks, they couldn’t say much more. One of them said the photos were too poor in quality to compare to the molds. A second wrote that the marks were “more likely than not made by insects or artifacts.” A third thought that some of the marks were probably bite marks, but he couldn’t match any of the molds to them. Two of the experts did indeed match one of the marks to one of the molds, but it was not Spence’s. It belonged to a housewife from Phillipsburg, Kansas. Unfortunately for Spence, the study wasn’t completed until after the deadline for Spence’s writ. He was eventually executed, despite numerous questions about his guilt—the biggest coming from the fact that the only physical evidence against him came from Campbell.
Campbell made at least one other embarrassing mistake that we know of. In 1984, a few years before he testified against Spence and Chaney, he was asked by a lieutenant in the sheriff’s department in Coconino County, Arizona, for help in identifying the body of a young woman found alongside I-40 near Flagstaff. The lieutenant had a hunch the girl was a missing runaway from Jacksonville, Florida, named Melody Cutlip, who had left home in 1981. Campbell compared the corpse’s teeth with those in a photo of Cutlip that he enlarged. “They matched exactly,” he told the Ocala Star-Banner. Cutlip’s family was notified and the corpse was buried in a Williams, Arizona, cemetery under a headstone with her name. In 1986, Cutlip contacted her mother. She was alive. Campbell was wrong.
A review of old bite mark cases will almost certainly reveal more false identifications, simply because of the nature of the way experts thought and testified. As Hales said in his affidavit in Chaney’s case:
At the time of the trial in December 1987 both the ABFO guidelines and the scientific field of Forensic Odontology supported use of the terms match and biter to relate a suspected person to a bite mark and it was permissible for experts to testify to a reasonable degree of medical/dental certainty that an individual was the biter in a case.
And indeed, if you go back and look at old cases, the word “match” is constantly used by experts, dating back to that very first 1974 case (“The bite mark matches the teeth reproduced in the model”). If experts didn’t say “match,” they said words that meant the same thing: “no question in my mind” (the defendant bit the victim); “it could be no one but [the defendant] that bit this girl’s arm.” Sometimes, as Hales did in Chaney’s trial, they would go further and use statistics, even though no studies had ever been done. Campbell did it in a 1977 Arizona case, when he testified that marks found on a murder victim’s breasts and a model he’d made of defendant’s teeth were “consistent,” which he then quantified by saying, “The probability factor of two sets of teeth being identical in a case similar to this is, approximately, eight in one million, or one in 125,000 people.”
Statements like these were, in Hales’s own contemporary words, “scientifically unsound”—opinions from well-intentioned experts with little to guide them but their own eyes and their own experience. (We reached out to Hales, who declined to comment for this article.) Campbell himself acknowledged the basic problem with bite mark analysis during the Joe Sidney Williams trial in 1987, when he was asked about its inherent subjectivity. “It is subjective,” he said. “I’ll admit it.”