What the Dog Knows (6 page)

Nor are there good comparative studies about the capacities of one animal's nose versus another, said Larry Meyers of Auburn University College of Veterinary Medicine. Repeat falsehoods enough, though, and they don't just take on the patina of truth, they become its replacement. Bears don't care, and inventions about their olfactory capacity probably matter less because bear handlers don't end up testifying in
criminal court about their abilities. Nor are bears' noses depended on for finding lost children alive.

Besides, there's just no great bear-nose grant money these days. The same isn't true for research on human noses. We're always interested in ourselves, and it's easier and safer to get human noses, rather than bear noses, into a laboratory. Even while we tend to denigrate it, the human sense of smell is getting more deserved attention. Humans can detect thousands of different odors. Even Linda Buck, Nobel olfactory biologist par excellence, hems and haws over some details: “It is estimated that humans, for example, can detect from 10,000 to over 100,000 different volatile compounds.” That's a difference of a factor of ten, less than some of the variations I've seen used with dogs, but hardly a rounding error.

While smell isn't entirely a lost subject for research or practice, for most Westerners, it is a deeply underappreciated sense compared to vision. It wasn't always so. Smell used to be a critical tool for physicians. Hundreds of years before we started exclaiming over the miracle of dogs being able to detect diabetes or lung cancer, doctors were using their noses to do the same thing. “Evaluating effluvia” was considered a basic diagnostic skill: Sweat on a rubella patient smelled like “freshly plucked feathers”; life-threatening diabetic ketoacidosis made a patient's breath smell like “rotting apples”; a certain bacterial condition made the skin smell like “over-ripe Camembert.” Now we leave that job to lab tests and litmus paper.

Vestiges of human scent skills do survive in pockets in the West. For instance, “odor mitigation” expert Larry Sunshine will fly into a city, tilt his head, open his nostrils, and identify specific foul odors in subways, including mildew and chemicals venting off plastic seats. Luca Turin, a perfume expert and biophysicist, can deconstruct a perfume down to its exact compounds and describe it in terms that make you want to laugh and weep and even buy that particular patchouli: “The smell was at once beguiling, salubrious, and toxic, and felt like a perfume composed for a fiercely intelligent librarian.”

Research is starting to blossom in this world of human olfaction, and it's taking some interesting turns—ones that working-dog handlers can relate to. Certain scents turn humans into the kinds of tracking machines whose accomplishments seem to rival those of trained scent dogs. If chocolate is involved. Scientists took a group of Berkeley undergraduates, showed them a video of canine scent-tracking, and then took them out on a lawn that had been laced with essential oil of chocolate. Scientists gave the students blindfolds, thick gloves, knee and elbow pads, and let them loose, off-lead. Could they track the chocolate using their noses, crawling on their hands and knees? Absolutely. The track they followed looked like the zigzag of a dog's trail.

The Berkeley scientists loved the students for more than their noses. Unlike rats or dogs, they could be debriefed afterward. The students could even say which nostril they tended to depend on: “Humans are an appealing animal model for addressing such questions because they can follow task instructions and accurately report behavioral strategies,” noted the article in
Nature Neuroscience
.

Fortunately, that's one aspect of olfaction that a few researchers are starting to focus on moving forward, both in humans and in dogs: the importance of training. Olfactory scientist Avery Gilbert has long argued that the trained human brain processes and understands scent differently. “I'd like to challenge my academic friends to stop giving random orders to college sophomores in the psychology labs,” he wrote, “and start observing odor fluency where it happens naturally—in creative people actively engaged in smell.”

In that respect, water is like chocolate. University of Pennsylvania researchers concluded in a study on people's ability to taste water that training, more than anything else, most enhances people's ability to identify smells. And during a repeat chocolate-tracking trial in Berkeley, students cut their tracking time by nearly two thirds. They learned to sniff faster. The students, taken out of the classroom and placed on grass, started to develop working noses.

•  •  •

Abalone and accelerants. Termites, truffles, and TNT. Crack cocaine and citrus cankers. Mildew, moths, melanoma. Peanut butter, pythons, and people. Spotted owls and spiny lobsters. Cows in heat. Gas leaks.

If there's a particular smell out there—illegal, endangered, delicious, destructive, invasive, or dangerous—handlers will try to train dogs to find it. The list of smells is lengthening daily.

Using hunting dogs who use their noses dates back many thousands of years. Human tracking and avalanche-rescue work originated centuries ago. But the explosion of tasks we've found for dogs in the past four decades often reflects our all-too-human tendencies toward violence, addiction, arson, and excess. We have entered the era of the sniffer dog.

Working-dog historians may argue about when that era began. The first handful of U.S. police-dog programs were developed at the turn of the twentieth century in New York City and New Jersey. Over the next few decades, units emerged in Connecticut, Berkeley, Pennsylvania, and Detroit. Fewer than twenty programs existed up through the early 1950s. After that, police K9 units reproduced rapidly.

Much of the expansion coincided with the military's experimental animal science that began in the mid-1960s and lasted into the early 1970s, during the Vietnam War, when researchers realized that sentry and scout dogs could be trained for additional sniffing tasks: finding punji pits and trip wires, not to mention illicit drugs in soldiers' barracks in Vietnam. On the domestic front, the military-research scientists speculated that dogs might be used to find bombs and weapons. Hijackings, bombings, and assassinations were on the rise in the United States: what one researcher called “the ills of the 1960s.”

It wasn't just in law enforcement that the uses of dogs' noses started multiplying. Seattle area trainer and handler Marcia Koenig, one of the early volunteer search-dog team members, tracks the history of volunteer
search-dog teams to 1962, when the German Shepherd Dog Club of Washington State formed the first air-scenting search-dog group. One of its founders, Bill Syrotuck, wrote the clear, concise book
Scent and the Scenting Dog
(one of the books Nancy Hook had permitted me to read). Today volunteer groups deploy to search in wilderness areas, in avalanches, on water, and in disasters. Marcia Koenig estimates that the U.S. has more than five hundred volunteer teams.

While working dogs were off and running by the mid-1960s, it's best not to feel too sentimental about what that growth represents. Each time a dog accomplishes a particular task for humans isn't automatically a moment for celebration. Dogs may have co-evolved with us, but they don't have a lot of say in how we decide to use them, so the “co” gives a false impression of equity. The dog mostly tries to please us, using its “canid tool kit of flexible sociality, a good nose, and expertise in hunting,” as John Bradshaw, Foundation Director of the Anthrozoology Institute at the University of Bristol, puts it.

Good working dogs have to move swiftly, hear acutely, smell well, and communicate clearly with their handlers—even bite on occasion. Since domestication, they've been used as adjuncts for the evil that people do, as well as the good, and sometimes both at once. They can be used to consolidate or pervert power in concrete ways. They can track a slave, a lost child, or a rapist without distinguishing. They can help suppress peaceful civil rights protesters or control an angry mob that's up to no good. People create the problems, and working dogs come along for the ride. Right now, as we engage in conflicts in the Middle East and South Asia, we use dogs to find bombs and control the groups fighting against us. When domestic boas and pythons get too big for Florida apartments, they are dumped in the Everglades, and we use dogs to locate them before the snakes decimate the native wildlife. We have huge prisons filled with contraband and cell phones, and we use dogs to find them.

Though we've been using working dogs for tens of thousands of years, academic researchers are just starting to catch up with what these dogs do and how. Characterizing and understanding dogs' olfactory
and cognitive skills isn't best done in a lab with limited and usually untrained dog subjects. If working dogs are overrated in the popular imagination, they have been mostly underrated in science, although that is rapidly changing. Nonetheless, far too much scientific work on dog cognition and olfaction has been done on pets—dogs who don't use their brains and noses for a living.

Just as Berkeley scientists took undergraduates out of the psych lab and onto the grass to track, just as Avery Gilbert urged neurobiologists into the kitchen of a great chef or the lab of a master perfumer to understand how human experts process scent, some working-dog experts are urging psychologists and neurobiologists to start using working dogs as research subjects.

The first scientific problem-solving tests with dogs appear to be those of Edward Thorndike, who served as president of the American Psychological Association at the turn of the twentieth century. He created “puzzle boxes” (precursors of the Skinner box), put domestic dogs and cats in them, and waited for them to figure out how to escape. The performance of both species was disappointing. Thorndike's conclusion was that operant conditioning, rather than any independent cognitive function, was the only thing that helped the beasts escape. The conclusion held sway in the research community for over a century. “Based on Thorndike's experiments and others like them, scientists now believe that dogs have rather limited powers of reasoning, certainly inferior to those of chimpanzees (and even a few birds),” John Bradshaw wrote in 2011.

And so it goes, even as dogs' noses become more and more popular with the public. For instance, one group of studies showed that dogs ranked from “okay” to “fine” at distinguishing between fraternal and even identical twins' scent on gauze pads. When dogs were asked to choose the scent difference between identical twins living in the same house and eating the same food, however, they failed miserably. That study's conclusions were clear about the outer limits of dogs' scenting capacity.

These are exactly the kinds of studies that irritate cognitive psychologist William “Deak” Helton, at the University of Canterbury, New Zealand, an expert in the science of working dogs. Experiments like the twin-distinguishing studies, he said, are akin to putting undergraduate students on a flight simulator, studying their performance, and coming to conclusions about the capabilities of trained pilots. One doesn't need to go all the way to the vastly experienced Captain Chesley “Sully” Sullenberger safely landing a US Airways jet on the Hudson River to think Deak might have a point.

“More studies need to be done on highly skilled dogs,” Deak said. “The problem, of course, is these dogs are already likely to be working and are too valuable.” It's not that the research done with untrained dogs is invalid, he noted; it's that those dogs haven't had their noses—or their cognitive abilities—trained and developed.

The twin-distinction story ends well. A group of scientists and ethologists in Czechoslovakia, where much of the groundbreaking work on dog cognition has taken place, decided to do another study on the identical-twin scent-discrimination problem. These researchers used trained scent-detection dogs. Their 2011 study showed that well-trained German shepherds could easily and correctly distinguish between the scent profiles of identical twins, even those who lived in the same environment and ate the same food.

Properly trained and handled, dogs will find almost anything we ask them to. For many substances, they can find small amounts. A 2006 study (albeit with a tiny sample) showed that trained dogs could detect one to two parts per trillion of n-Amyl acetate, a banana-scented solvent. That's the equivalent of a drop of water in twenty Olympicsized swimming pools.

It's not universally true that dogs can smell at much lower concentrations than humans do. Larry Myers, who has been doing research on sniffer dogs since 1982, thinks it's silly to try to quantify which dogs have the best noses, which species have the best noses, or even to compare human noses with dog noses. Yet he couldn't help doing a couple
of quick, informal experiments comparing his lab workers' noses with dogs' noses on odors. Myers tested one group of workers and dogs on acetone, a cleaning solvent common around labs. “My lab workers could smell it at lower concentrations than the dogs could.” But with eugonal, a carnation-ey, clove-y compound that Myers started using as a standard pretest compound before having dogs perform more elaborate tests, they showed a response at one millionth of the concentration that lab workers could.

The old dogs' tales don't end at levels of concentration. There's also the strict correlation between the number of dogs' smell receptor cells and their scenting ability. At first, as I read the nose literature, I ranked a German shepherd's nose below a bloodhound's but well above most other breeds'. Solo's nose on this spreadsheet was very good but not excellent. Much of the best-selling dog literature promotes the concept that the more smell receptor cells a dog has in its nose, the better its scenting ability: “For example, the Dachshund has around 125 million smell receptor cells, while a Fox terrier has 147 million and the German Shepherd has about 225 million.” The bloodhound has all those breeds beat, with 300 million receptors.