The idea of using teeth to identify someone has a long history. One of the earliest recorded cases (and perhaps the very first one) dates back to the first century. The wife of Roman Emperor Claudius, Julia Agrippina, commanded a colonel of the Roman guard to order the suicide of her rival, a former empress and consort of Caligula. As proof of the deed, Agrippina wanted her head brought back for examination. Unfortunately, she could not recognize the features of the decayed head, but she did remember that the woman had had a unique set of teeth. It was by looking inside the mouth of her rival that Agrippina was satisfied. The deed had indeed been done.
The identification of remains based on examining the teeth is one part of the discipline known as forensic odontology, which is the application of dentistry to the field of law. Forensic odontology also concerns itself with interpreting oral injuries, opining in cases of dental malpractice and, most controversial of all, comparing bite marks.
Bite marks can be seen in cases of homicide, child abuse, and sexual assault. Bite mark comparisons can be done in these cases by specialized dentists. If the marks left on a victim can be matched with certainty to a particular suspect, a dangerous criminal can be put behind bars. This bit of pattern matching makes intuitive sense. Our dentition is probably unique, we think to ourselves, and if you can measure it all—the dimensions of teeth, the space between them, how crooked they may be—you may end up with the oral version of a fingerprint.
But twenty years of assessment by expert committees have revealed a troubling fact: bite mark comparisons are built on a very shaky foundation, riddled with holes and propped up by anecdotes. Even though jury-eligible people may believe the odds of a false match for bite marks to be one in a million, we now know it is orders of magnitude higher.
The story of forensic bite mark comparisons also highlights the inconvenient fact that, although science is self-correcting, scientists are not always happy to admit they were wrong.
A cheesed-off criminal
In 1954, the first legally published case of bite mark comparison in the United States was presented and it involved cheese. It was a burglary and a piece of cheese that was found at the crime scene, complete with teeth indentations. Could those teeth marks confirm the guilt of the man who was eventually captured? The suspect was made to bite into a fresh piece of cheese and the two marks were compared not by a dentist but by a firearms examiner, who concluded they matched. A dentist later testified at trial and confirmed the match. Bite mark matching had now made its way into the American court system, and though it would remain rare for a couple of decades, it would eventually join the ranks of a respected body of forensic sciences and be used to eliminate suspects and reinforce the case against others.
But in the mid-2000s, some scientists took a bite out of the reality sandwich and claimed that this forensic method with the power to confine people to jail was built on an illusion of objectivity.
More holes than Swiss cheese
“The forensic science system, encompassing both research and practice, has serious problems.” This is the arresting take-home message of the preface to a long report by the National Academy of Sciences published in 2009. As it turns out, many forensic tests, especially those involving comparison between patterns, were never rigorously tested by scientific means, and that, the report indicated, included bite mark comparisons.
For bite mark comparisons to lay on a scientific foundation, there are two big assumptions that need to be true. The first is that my teeth and your teeth are different. It’s the assumption that our dentition is unique, specifically in terms of where the teeth involved in biting are placed in the mouth. While this is widely believed to be the case even by experts in the field, the evidence for this is very, very thin and comes from less than a handful of problematic studies. This gap between belief and reality came to a head in a 2011 paper. Scientists examined 344 different human dental casts, measured them very precisely using 3D laser scanning, and compared them. A theoretical model predicted the probability of finding even one match in this set based on five teeth would be less than one in a million, but the precise measurements and comparisons yielded 32 matches. We can expect that, when measurements are much less precise in real-world scenarios, the number of matches would go up.
The second assumption underlying the discipline is that whatever is bitten—and in criminal cases, it is often human skin—will allow for the (alleged) unique bite mark to be preserved. But a bite mark in skin is not like a bite mark in plaster. Skin is a living and elastic medium. The shape and size of the muscles and fat reserves underneath, the skin’s curvature, its looseness, the presence of fluid underneath, its recoil, the movements of the victim during the attack, the subsequent bruising—all of these factors affect how teeth marks will be preserved in the skin for later examination. In fact, the National Academy of Sciences report points out that when bite marks are created in the lab, the skin deforms them so much and with such variation that comparing bite marks cannot reliably exclude or include a suspect as a potential biter.
This blunt report paved the way for the President’s Council of Advisors on Science and Technology, whose 2016 report to President Barack Obama on forensic science was scathing. The council examined the evidence behind bite mark comparisons and concluded that it was “far from meeting the scientific standards for foundational validity.” Only a small percentage of the papers published in the scientific literature about the technique were actual experiments as opposed to case reports. The prospects for a genuine science to emerge from these ashes was qualified by the report as low. “We advise against devoting significant resources to such efforts,” the council wrote.
Correcting course is like pulling teeth
The way in which bite mark comparisons were enshrined as science helps illuminate why, despite damning expert reports, the practice is still used in courts. It has been argued that forensic odontologists once agreed that you could not look at a crime scene bite mark and accurately identify its source. The 1954 case of the cheese-eating burglar was a bit of an exception. An even more note-worthy exception was the case in 1975 of People v Marx, where a murdered woman was bitten on the nose, and the exception ended up becoming the rule. “Ironically,” a 2016 paper denouncing the exaggerated claims of forensic bite mark identification reads, “rather than forensic dentists convincing courts that their field could accurately identify the sources of bite marks, the courts convinced forensic dentists that they could do what until then they doubted they could do.” Because the legal system is built on precedent, kicking bite mark comparisons off their scientific pedestal did not remove them from legal consideration. If bite marks were used in the past, they can, it is argued, continue to be used now. The legal system does not always deal well with the evolving body of knowledge in the sciences. It has a built-in argument from antiquity: if it’s been used for a long time, the faulty argument goes, it must be good enough.
It is becoming harder and harder to resist labelling bite mark comparison as a pseudoscience, yet some forensic odontologists criticize the pushback against this technique. One often quoted argument is that making these comparisons ineligible in court would hamper efforts to convict some dangerous criminals. By that logic, not allowing palm readers to testify in court carries the same risk.
When they have been tested for their expertise, forensic odontologists comparing bite mark patterns have not performed well. A 1975 study showed one in four incorrect matches when using perfect conditions, with that number climbing to nine in ten when bite marks were photographed 24 hours after having been made. At a 1999 workshop for forensic odontologists, participants were asked to match four bite marks to seven models, and while the exercise was retrospectively denounced as not having been a formal proficiency test, nearly two in three matches reported were wrong. A 2001 study estimated the rate of false matches using the technique as being between 11.9 and 22%. Worse: forensic odontologists cannot reliably agree if a mark was left by human teeth or not.
This pseudo-expertise can have grave consequences. The Innocence Project, dedicated to exonerating the wrongly convicted through DNA testing, reports that at least 26 people in the United States have been wrongfully convicted, arrested, or charged based on bite mark comparisons. Once the pseudoscience genie is out of the bottle, it’s very hard to put it back in.
- Bite mark comparisons are one part of forensic odontology, a practice that puts dentistry in the service of the law
- The assumptions that human dentition is unique and that bite marks left in skin can lead to useful and reliable matches are not supported by good scientific evidence
- When dentists who compare bite marks for legal cases are tested, they tend not to agree with each other on whether the marks match a suspect and even on whether the marks were left by a human being