by Cat Warren
Why bother sorting out the chemicals from that messy pile? Because ultimately, the knowledge may help create a machine that can help detect the smell of death. It may help scientists develop more effective pseudoscents for dogs to train on.
If it’s a new frontier to figure out what compounds are in the air column from human decomposition, it’s another stratospheric leap to understand what the dog’s nose is picking up from that air column and translating as human remains. No one knows exactly what the dogs are smelling. We can’t ask them. Most likely, they are smelling a lot of things mixed together.
“It’s a much more complex chemical composition than any other forensic sample, with the possible exception of human scent,” said Florida International University analytical chemist Kenneth Furton. He and a group of scientists, trainers, and law enforcement representatives are trying to develop a national set of best practices for detection dogs. Creating a set of best practices is a challenge, and understanding how best to train dogs to detect human decomposition is one of the group’s biggest challenges. “There are more knowledge gaps in human-remains detection than anywhere,” Furton said.
The things that we humans invented—bombs, manufactured drugs, and land mines—are chemically simple in comparison to our remains. Decomposing humans are not entirely a black hole, though. While no one has yet nailed the perfect formula for what it is that good cadaver dogs think is unique or interesting about dead people—as compared with, for instance, garbage from a suburban household—we do know that solidly trained dogs can tell the difference between dead human, dead deer, and aged goat cheese, or something equally putrefied and odiferous. Arpad noted that cadaver dogs tend to alert on dead sheep more than on other species; he thinks some dogs get confused by the generous dose of sulfur that sheep emit after they die. We emit it, too.
Although sheep may be chemically close to humans, there are, he notes, several significant chemical differences between us. Chemicals we ingest may play a role. Arpad’s research laboratory, along with one or two others, found tantalizing evidence that compounds we swallow or inhale—from fluorinated water to asthma inhalants—may play a role in our unique smell after death. It’s not much of a stretch to think that our decaying bodies might whisper more than a hint of je ne sais quoi of chemical compounds we’ve steeped our bodies in during our lifetimes.
“We take in a lot more chemicals than we should,” Arpad said. The unanswered question is whether those translate to volatile compounds significant to the dog, even though the carcinogenic compound carbon tetrachloride seems to jump out when Arpad takes air measurements of human decomposition.
We have a fair amount of evidence that dogs do fine at detecting remains that are hundreds of years old—long before fluorocarbons and freons and fluorinated compounds and solvents and poisonous cleaners and antibiotics came on the scene. In rural areas where water isn’t fluorinated, dogs can find people who have mostly drunk well water all their lives. Domesticated pets ingest plenty of fluorinated city water and chemicals, and well-trained dogs don’t alert on their remains.
Arpad believes they are getting close to knowing what dogs find important and unique about human remains. The chemical portrait gets blurrier with buried bodies, which may be emitting different volatiles than surface ones. He thinks he knows what volatiles dogs are alerting on in burials—perhaps as few as thirty compounds—but he’s not yet tested the theory. Those volatile gases have to be available at the soil surface for the dog’s nose to collect. Even bone has twelve detectable volatile compounds.
It is possible that a unique volatile compound—something that barely registers on the gas chromatographs positioned around bodies in Arpad and his colleagues’ experimental research plots—hits the dog’s nose and lights up her brain like a pinball machine. Or it could be a few compounds, or a bunch of them in various delightful combinations. When you combine the variety of conditions under which dogs find human remains, from freshly deceased to hundreds of years old, the options become dizzying. For instance, Arpad notes, the odor profile for a body that has all the busy microbes associated with the digestive system working overtime is very different from the profile for scattered limbs.
While we don’t know the exact compounds they are smelling, dogs find human remains. That should count for something. Just because dogs find the dead without an utterly coherent scientific theory as to why, and without an easy way to test it under controls, doesn’t mean it’s not happening.
Cadaver dogs’ ability to find human remains may be analogous to humans’ ability to recognize faces. Researchers don’t have a clear understanding of how facial recognition works, but people do it effortlessly, at all angles, even in bad light. Trying to break that recognition down into parts (it’s those great cheekbones!) doesn’t work. It’s the whole face. Machines still aren’t nearly as adept at facial recognition.
So the two scientific arenas—what volatiles in what percentages are involved in human decomposition itself, versus what volatiles dogs react to as cadaver scent—may overlap a great deal. Or not. Who cares? Well, we like to know things. It might help us figure out why some dogs are drawn to rotting trees like sycamores or to the vegetative decomposition in swamps. Isolating what the dogs find significant in decomposing humans might help handlers and trainers identify which training aids are best and the ideal way to store them. Ultimately, that knowledge might lead to finding substitute training aids that are legal, safe, and much closer to the real thing than the pseudoscents or other chemicals in current use.
This new research arena has started to upset old beliefs. Some trainers and handlers—more in the past but a few in the present—have claimed that pig samples are the next best thing to human samples for cadaver-dog training. The temptation is great. Pigs are plentiful and a forensic fallback, a model for studying human disease and decay. No ethical fuss or muss about using them as training aids: Just pick up some pork spareribs from the grocery meat counter. Yet, as Arpad Vass noted, there’s a vast difference between the chemical profiles of deceased humans and swine. “Pigs smell quite different from humans. We have this well documented.”
Nancy Hook scoffed from the beginning of training about handlers using pig tissue as training samples. It was pretty simple to her: “Pigs aren’t people.”
Mary Cablk, an associate professor at the Desert Research Institute in Nevada who trains cadaver dogs and does research on the reliability of detector dogs, took the pig problem a step further. She and her analytical chemist husband, John Sagebiel at University of Nevada, Reno, compared the volatile organic compounds of chicken, cow, and pig with those from human remains. Their results should be the final nail in the coffin for an entire cadre of trainers and handlers who have said for decades that pig samples are the next best thing to human samples for training. Their research shows that we humans smell much more like chicken than pig when we decompose. If it’s any comfort, we smell like organic chicken from Whole Foods.
• • •
. . . and suddenly, coincident with death the beam end dropped with an audible stroke hitting against the lower limiting bar. . . . The loss was ascertained to be three-fourths of an ounce.
—Duncan McDougall, MD, “Hypothesis Concerning Soul Substance,” Journal of the American Society for Psychical Research, 1907
Cadaver-dog handlers and trainers have watched dogs find human remains for several decades. No matter. No rigorous scientific studies had shown how well dogs could do it. They just did it: Dogs were indicating that they smelled human decomposition. No one knew exactly what they were alerting on, unless a body, or a part of a body, was there as undeniable proof. What part of the part, though, were they alerting on? How soon after death could dogs detect something? How long after death did the scent last? And what, for a dog, was that “scent”?
Despite numerous studies of how working dogs reliably detect drugs and bombs, few scientific studies have featured cadaver dogs. It’s messy work, and the real-world aspect can irrita
te scientists in search of conclusive data. It’s hard enough figuring out how a narcotics dog can find pure heroin; figuring out how cadaver dogs can find the infinite variables of dead humans seems impossible.
Scientific uncertainty hasn’t kept people from using dogs worldwide to help pinpoint the perfume of death, from the faintest emanation off a tooth to the fulsome scent that lofts from an entire body. A few small studies existed here and there, like Debra Komar’s 1999 study on cadaver dogs’ ability to find scattered human remains in Canada, and another in 2003 on dogs’ finding buried remains in the southeastern United States. That 2003 study noted quite accurately that “dog handlers affected the reliability of the cadaver dog results.”
It was far past time, at the beginning of the twenty-first century, for someone to come up with a controlled scientific study on cadaver dogs’ capacity to detect the faintest smell of death. That was when an unfortunate series of events—almost always the case when cadaver dogs are involved—created a perfect opportunity for imagining, then realizing, a small, elegant scientific study.
A wife in Hamburg, Germany, disappeared off a yacht in 2005 while sailing, apparently not entirely happily, with her husband. The husband reported her missing. She must have fallen overboard, he said. When the Hamburg state police sent one of their cadaver dogs onto the boat, the Belgian Malinois aggressively communicated to its handler that something bad had happened on the mattress in the yacht’s bedroom.
There was no body in the bed. No blood. No tissue. It must have been a soupçon of scent that caught the Malinois’s nose. Who knows? Mattresses can be funky places, even on the best yacht. Regardless, without a body, tissue, or blood, there was no case.
The prosecutor, however, wanted to believe the dog. He contacted Lars Oesterhelweg, a forensic pathologist then at the Institute of Legal Medicine in Hamburg, and asked him to set up a study—not replicating the entire yacht, but providing more definitive proof that dogs can scent death without the presence of specific forensic evidence, like blood, tissue, or bone.
For his study, Oesterhelweg and his colleagues used two recently deceased men, A and B, sixty and sixty-three years old. They had collapsed and died on the streets of Hamburg. Sometime before, they had agreed to donate their bodies to medical science. A and B could not have known how delightfully and noninvasively their bodies would be used. Each freshly dead man was whisked into the local hospital’s inner courtyard, wrapped in a cotton blanket, and laid, for just two minutes and ten minutes each, on top of brand-new carpet squares that sat on new tables in the open-air courtyard. Oesterhelweg didn’t want a hint of hospital contamination. The carpet squares received nothing but that indirect exposure to two-hours-old bodies: no tissue, no fluid, no blood, no rafting cells, no DNA. My romantic, noninterventionist side likes to imagine that was the beginning and end of A’s and B’s sacrifice to medical science: this gentle wrapping, this brief application of their bodies to carpet squares. But that’s neither the reality of good cadaver use nor good recycling.
For the next two months, three Hamburg police cadaver dogs were asked to show their handlers which carpet square, among a group of uncontaminated carpet squares, contained cadaver scent. The most mysterious fragrance on earth was no mystery to the dogs. Two of the dogs, B and L, were almost 100 percent accurate. K scored 90 percent. That’s the reality of working dogs. A few are excellent, while some are very good. (Others are horrid. The last category wasn’t represented in the study.) The small study’s results: Well-trained cadaver dogs can smell the faint remnants of the odor of death, impregnated on a carpet swatch, for months following the brief presence of a newly dead person.
The study showed, Oesterhelweg wrote in what amounted to scientific effusion, that well-trained cadaver dogs are “an outstanding tool for law enforcement.” I tried to find out, to no avail, if the prosecutor had pursued the case. Never before had I wanted so much to be able to communicate in German.
Every study has its limitations. This one didn’t answer the question: What exactly were the dogs smelling in those carpet squares? They could alert reliably, according to the study, but what was in that fresh perfume of death?
Arpad wistfully asked me if the German forensic scientists had done a headspace study on that early perfume. Did they measure what volatiles were in the airspace in the containers that held the carpet? I doubted it, but I could see why Arpad wanted the information. By the time donated bodies get shipped to the University of Tennessee Anthropological Research Facility, their decomposition is much further advanced. In the seconds and minutes after death, Arpad thinks it’s possible that compounds such as ammonia, hydrogen, carbon dioxide, and methane escape from the body. Such compounds don’t weigh much at all.
They are lighter than air.
• • •
People are spending more time worrying about the possible nuances of training than going out and training their damn dog. Our dogs are very forgiving, so you want to try working on some source. See what your dog does. It’s not a big frigging mystery.
—Andy Rebmann, 2012
Every training aid for every working dog has its drawbacks. If you are training a bed-bug dog, you must find a way to keep your bed bugs fed and happy (and contained). If you are training a bomb dog, you have to be comfortable carrying nitroglycerin and gunpowder around in your trunk.
If you have a cadaver dog, you run into a different kind of supply and storage problem. Having diverse materials to train on is crucial, and “decomp,” as it’s called, comes in all varieties: from recent blood on a surgical swab, to lovingly harvested dirt from a Civil War burial ground, to six-hundred-year-old bones from the Mississippi Delta.
Happily for Solo, North Carolina has reasonable laws about cadaver-dog training material. The wisdom tooth of Nancy’s ex-husband was only the beginning. Solo would need a variety of materials to train on, from fresh to older to ancient. Material that would, albeit at a scaled-back level, mimic what he might find out in the woods, buried, or in an abandoned building. So in the midst of answering student e-mails in my university office, I’d find myself distracted, mail-ordering a search-and-rescue dog package from the Bone Room in Berkeley, California—although I called someone at the state medical examiner’s office first to make sure it was legal. It was. I was delighted when the order arrived in the mail: a small box with an ivory tibia and some toe bones in a plastic bag. The foot bones looked remarkably like the Styrofoam peanuts they came nestled in.
Less than a year later, evidence started mounting that some of the plastinated bodies displayed in the popular museum show Bodies: The Exhibition came from prisons in China. Curious, I started researching the possible source of my specimens from the Bone Room. The results were inconclusive, but I realized that I needed to find other sources. So, obviously, did the Bone Room, which posted on its website: “We regret to inform you that our stock of damaged and discolored bones, set aside as Search & Rescue Dog training packages, has finally been totally decimated, and we will not be able to offer them for a while. Unfortunately there are absolutely no human bones coming into the country at this time, and while we are currently trying to find a source that would hopefully include more of the ‘dog quality’ specimens, we do not yet have a line on a possible exporter.”
When having my teeth cleaned, I talked with my dentist about how to get hold of a couple of molars for training. He was happy to provide a few, and kind enough to lend me, briefly, the jawbone from an anatomical teaching skull that had been in his office for decades, to see how Solo would behave. Solo reacted, though he was clearly surprised when he put his nose next to it. He was underwhelmed by the source: That’s it? I reacted to that bit of scent? I, on the other hand, was thrilled and reported the outcome to the dentist when I returned the jawbone.
A K9 officer from a nearby city provided me with carpet from a processed suicide scene. One warm summer evening found me with a friendly death investigator who had handled cadaver dogs. We were in the parking lot of the
police K9 unit’s headquarters, cutting up material for training aids: an old sheet that had lain for days underneath an undiscovered body in an apartment. It was malodorous but not nauseating, and we didn’t need the whole thing: just six-inch-by-six-inch patches that we could pop into Mason jars. The scissors would be bagged and tossed. We knelt in front of our bounty to make careful cuts, then rocked back on our heels. We had to make sure our double-nitrile-gloved hands didn’t touch anything except the material. We were smiling. Solo was whining in the car, waiting for training. It was a beautiful evening.
I permanently borrowed a wide variety of kitchen storage containers from David, whose home-chef habits benefited me on several fronts: Mason jars, honey and jam jars, bigger Rubbermaid and Cambro and gallon glass jars for holding the smaller jars, all storage for Solo’s training materials. I bought others new. Using an assortment would prevent Solo from associating the smell of any one container with the scent of the dead. If I were ever called to testify, I wasn’t going to be trapped by a good attorney into admitting it was possible that my dog was trained to smell and alert on plastic storage containers, or Ziploc bags, or cotton sheets, instead of the human remains I was trying to train him on.
There’s great debate about whether dogs get rewarded accidentally for finding whatever storage material you are using instead of the training material you want him to find. Or for alerting on the smell of the gloves you use to place the samples, for that matter. Some trainers and handlers contend that material should be stored only in glass jars with metal lids, or it will be contaminated, and you will inadvertently cross-train your dog on plastic bags. Others insist that material needs to be stored in everything and everywhere: teeth in the freezer in a freezer bag, dried blood in the fridge in a jar, bone in the pantry.
Almost everyone believes that if your state laws allow it, it’s ideal to train your dog on the wide range of stages humans go through before they disappear entirely—from fresh tissue and blood, to what’s called greasy or wet bone, to adipocere, to the dry bones of the desert, and even the ashes of the crematorium.