Any time Alexandra Horowitz releases a new book is cause for celebration here at Bark. We’ve been fans since her 2009 hit, Inside of a Dog, and have continued to follow her work as she uncovers new insights into our co-pilots’ internal lives and external behaviors.
In addition to teaching psychology, canine cognition and creative non-fiction at Barnard College, Columbia University, she also leads the college’s Dog Cognition Lab.
Her new book, Being a Dog, delves deeply into the primacy of dogs’ sense of smell, and we talk with her about what she found.
Bark: Has anyone studied why some dogs are better at smelling than others —is it genetic or is it drive?
Alexandra Horowitz: Everything I’ve seen points to drive being the major indicator of whether a dog will be good as a detection dog: drive to find the odor, to keep working when frustrated, to get to the reward (like a game with a tug toy) at the episode’s end.
This is not to say that breed is irrelevant: some breeds are naturally more driven to pursue an odor relentlessly, or are driven to do whatever it takes to get a game of toss with a tennis ball. And some dogs—like Bloodhounds and Beagles—have more olfactory cells in their noses and more equipment around their faces (long ears, drooly jowls) to help bring odors up the nose. They may smell odors at lower levels.
Curiously, though, the notion that certain breeds are inevitably better at detection work than others hasn’t been borne out. It’s tradition more than science.
B: I was once told by a woman who handles tracking Coon Hounds that dogs can show a preference for how they scent; talking about the same breed, she said some sniff the ground, while others prefer sniffing the air. Have you observed individual differences in the same breed in your research?
AH: Absolutely. Different dogs have different sniffing tactics; “on the air” or “on the ground” are the two ways dogs try to pursue a scent. Often, though, these are distinguished by task, not by dog—that is, if a dog is tracking a distant (old) scent, on the ground makes more sense; the odor is probably no longer in the air. But a dog trying to locate someone/something who has recently passed by will be air-scenting.
B: Can adult dogs can identify their littermates or their mother by smell?
AH: In theory, this would be trivially easy for dogs. All dogs have their own “signature scents” (as do we, to dogs), so there would be no trouble distinguishing dogs of one’s litter from other dogs. Now, the question of whether an adult dog who has been separated for years from her littermates/family can recognize them is a different question: it’s more about memory than about perceptual ability. Memory is fallible in humans, and it is fallible in dogs. We forget. So it’s quite possible that, even having known one’s family by scent, it would be later forgotten. (But there is also good reason to believe that a trace would remain—that distant memory one cannot quite place.)
B: While you note in your new book that puppies at the Working Dog Center at the University of Pennsylvania aren’t formally trained until a certain age, they do receive some kind of training, right?
AH: Yes, they are being “trained” to some degree: I think Dr. Otto and the Penn Working Dog Center trainers would agree when I say they are being trained to be good general-purpose working dogs. As I describe in my book, I saw dogs being put through their paces in lots of different (what were to them) games: find the missing person, find the hidden scent. They are being exposed to unusual sounds and environments and getting acclimated to them. They are learning the skills of detecting something, working with someone, and loving it. And they do.
B: Do working dogs get nose fatigue— do they reach a point at which they can no longer reliably follow a scent? If so, what do the pros do to work around that?
AH: The phenomenon of the nose no longer noticing an odor—adaptation —happens to us within minutes. Walk into a coffee shop, take in its familiar odors and a few minutes later, you might smell … almost nothing. The receptor cells in the nose that noticed the odor simply stopped responding after continued exposure.
The cells in the dog’s nose work similarly, but any dog employed as a detection dog is doing something different. Because they continue to sniff different areas of the odor “scene,” their noses won’t turn off to the smell. Tracking dogs are also known to simply lift their noses from the ground once in a while and sniff the air, as though to clear their noses.
On the other hand, working dogs certainly get fatigued from too much stimulation and too much exertion. Handlers know their dogs and will read their dogs’ responses to know when they need a break.
B: In the book K9 Scent Training by Resi Gerritsen and Ruud Haak (leading specialists in identification, tracking and detection-dogs), I read that female dogs are better at smelling than males. Did this also come up in your research? Any idea if the same can be said for our species?
AH: Gerritsen and Haak are great resources on detection-dog training and skills. I suspect their assessment comes directly from their own and other trainers’ experience with dogs. I don’t doubt it, though I don’t believe that the question has been formally tested. Interestingly, women are often said to be “better smellers” than men, and research does bear this out (on average, of course).
B: I’m quite curious about the canine visual sense vis-à-vis their olfactory sense, especially for dogs of the sporting breed. When our Wirehaired Pointer is out in the field, she seems to rely primarily on her sense of smell; sometimes a rabbit’s been sitting just a few feet from her, but she doesn’t see it, or even seem to smell it. Is it “I can’t smell it so I don’t see it”?
AH: As with us, dogs’ senses work together. Only for dogs, olfaction takes priority. From that point of view, you can imagine how vision might aid smelling: if a dog detects an odor on the breeze, she can then look up and try to locate, with her eyes, the source of that odor (and then head toward it for closer sniffing!). When I watched the dogs at the Penn Working Dog Center do a “person search” for people hidden in large PVC barrels in a large field, the dogs used vision to guide them while smelling: first, they followed their eyes to head toward the barrels, then followed their noses to identify which one held a person.
A dog who is sniffing in the grass to a hidden ball (or rabbit) that is perfectly “visible” to someone else nearby is simply using olfaction first. By sniffing in the whole area around the hidden object, she creates an on-the-fl y map of where the object is; the closer she gets, the stronger the odor is. Sometimes, dogs rely on that much longer than we would expect before bringing fuller attention to what they see to aid their search.
B: I’m trying to train one of my dogs, Charlie, to find the poop of his housemate Kit while we’re out in the park; he’s actually pretty good at it. I started doing this after I noticed that he likes to pee on her fresh poop, and only on hers. How would you recommend I boost his proficiency level? And why the peeing on it?
AH: “Find poop!” Great game. And lots of dogs would be pretty good at it. Since Charlie started doing this behavior on his own, clearly little shaping was needed. The only task is pairing it with a request (like “find poop”) and making him aware that what he’s doing—which to him is “following that interesting smell”—is something that’s also valuable to you, so he’ll do it whenever you ask. If he’s not doing it reliably, then he doesn’t see its value to you. Better rewards! More reliable rewards! (But you and every good behavior reader know that.) And, taking a cue from working-dog handlers, you could pair an “alert” behavior—sitting, barking and so forth—so that he tells you when he’s found it.
What I learned from Sam Wasser, who trains dogs to find wildlife scat, is that what’s often difficult in training in the field (and you’re always “in the field”!) is to know yourself if the dog has alerted on the right scat. Once they are confident of their dogs’ alert, and don’t accept partial alerts, handlers can reward only for the correct scat.
As for his peeing on the poo, that’s a question I don’t think science has directly tackled. But we know that marking isn’t territorial in dogs; it seems to be information-leaving. It could be that a nice pile of stinky poo is a good place to leave your own mark.
B: Besides enrolling our dogs in nosework classes, what do you recommend that we do to tap into their world of smell and enrich their lives?
AH: Let them smell. If you live with a dog, start thinking about what the world is like from an olfactory point of view. Let them smell you (you are your scent, to your dog), let them smell each other (that’s how they find out who it is), and let them smell the world. Take walks for smelling (not just for peeing, or for exercise). The pleasure that comes from watching a dog snuffling down a path, nose to the ground and nose in the air, guided by nothing more than the filaments of odors that come his way, is to me unmatched.