Inside of a Dog: What Dogs See, Smell, and Know

For five minutes these dogs tumble, grab, bite, and lunge at each other. The wolfhound throws himself onto his side and the little dog responds with attacks to his face, belly, and paws. A swipe by the hound sends the Chihuahua scurrying backward, and she timidly sidesteps out of his reach. The hound barks, jumps up, and arrives back on his feet with a thud. At this, the Chihuahua races toward one of those feet and bites it, hard. They are in mid-embrace—the hound with his mouth surrounding the body of the Chihuahua, the Chihuahua kicking back at the hound’s face—when an owner snaps a leash on the hound’s collar and pulls him upright and away. The Chihuahua rights herself, looks after them, barks once, and trots back to her owner.

These dogs are so incommensurable with each other that they may as well be different species. The ease of play between them always puzzled me. The wolfhound bit, mouthed, and charged at the Chihuahua; yet the little dog responded not with fright but in kind. What explains their ability to play together? Why doesn’t the hound see the Chihuahua as prey? Why doesn’t the Chihuahua see the wolfhound as predator? The answer turns out to have nothing to do with the Chihuahua’s delusion of canine grandeur or the hound’s lack of predatory drive. Neither is it simply hardwired instinct taking over.

There are two ways to learn how play works—and what playing dogs are thinking, perceiving, and saying: be born as a dog, or spend a lot of time carefully observing dogs. The former was unavailable to me. Come along as I describe what I’ve learned by watching.

I am a dog person.

My home has always had a dog in it. My affinity for dogs began with our family dog, Aster, with his blue eyes, lopped tail, and nighttime neighborhood ramblings that often left me up late, wearing pajamas and worry, waiting for his midnight return. I long mourned the death of Heidi, a springer spaniel who ran with excitement—my childhood imagination had her tongue trailing out of the side of her mouth and her long ears blown back with the happy vigor of her run—right under a car’s tires on the state highway near our home. As a college student, I gazed with admiration and affection at an adopted chow mix Beckett as she stoically watched me leave for the day.

And now at my feet lies the warm, curly, panting form of Pumpernickel—Pump—a mutt who has lived with me for all of her sixteen years and through all of my adulthood. I have begun every one of my days in five states, five years of graduate school, and four jobs with her tail-thumping greeting when she hears me stir in the morning. As anyone who considers himself a dog person will recognize, I cannot imagine my life without this dog.

I am a dog person, a lover of dogs. I am also a scientist.

I study animal behavior. Professionally, I am wary of anthropomorphizing animals, attributing to them the feelings, thoughts, and desires that we use to describe ourselves. In learning how to study the behavior of animals, I was taught and adhered to the scientist’s code for describing actions: be objective; do not explain a behavior by appeal to a mental process when explanation by simpler processes will do; a phenomenon that is not publicly observable and confirmable is not the stuff of science. These days, as a professor of animal behavior, comparative cognition, and psychology, I teach from masterful texts that deal in quantifiable fact. They describe everything from hormonal and genetic explanations for the social behavior of animals, to conditioned responses, fixed action patterns, and optimal foraging rates, in the same steady, objective tone.

And yet.

Most of the questions my students have about animals remain quietly unanswered in these texts. At conferences where I have presented my research, other academics inevitably direct the postlecture conversations to their own experiences with their pets. And I still have the same questions I’d always had about my own dog—and no sudden rush of answers. Science, as practiced and reified in texts, rarely addresses our experiences of living with and attempting to understand the minds of our animals.

In my first years of graduate school, when I began studying the science of the mind, with a special interest in the minds of non-human animals, it never occurred to me to study dogs. Dogs seemed so familiar, so understood. There is nothing to be learned from dogs, colleagues claimed: dogs are simple, happy creatures whom we need to train and feed and love, and that is all there is to them. There is no data in dogs. That was the conventional wisdom among scientists. My dissertation advisor studied, respectably, baboons: primates are the animals of choice in the field of animal cognition. The assumption is that the likeliest place to find skills and cognition approaching our own is in our primate brethren. That was, and remains, the prevailing view of behavioral scientists. Worse still, dog owners seemed to have already covered the territory of theorizing about the dog mind, and their theories were generated from anecdotes and misapplied anthropomorphisms. The very notion of the mind of a dog was tainted.

And yet.

I spent many recreational hours during my years of graduate school in California in the local dog parks and beaches with Pumpernickel. At the time I was in training as an ethologist, a scientist of animal behavior. I joined two research groups observing highly social creatures: the white rhinoceros at the Wild Animal Park in Escondido, and the bonobos (pygmy chimpanzees) at the Park and the San Diego Zoo. I learned the science of careful observations, data gathering, and statistical analysis. Over time, this way of looking began seeping into those recreational hours at the dog parks. Suddenly the dogs, with their fluent travel between their own social world and that of people, became entirely unfamiliar: I stopped seeing their behavior as simple and understood.

Where I once saw and smiled at play between Pumpernickel and the local bull terrier, I now saw a complex dance requiring mutual cooperation, split-second communications, and assessment of each other’s abilities and desires. The slightest turn of a head or the point of a nose now seemed directed, meaningful. I saw dogs whose owners did not understand a single thing their dogs were doing; I saw dogs too clever for their playmates; I saw people misreading canine requests as confusion and delight as aggression. I began bringing a video camera with us and taping our outings at the parks. At home I watched the tapes of dogs playing with dogs, of people ball- and Frisbee-tossing to their dogs—tapes of chasing, fighting, petting, running, barking. With new sensitivity to the possible richness of social interactions in an entirely non-linguistic world, all of these once ordinary activities now seemed to me to be an untapped font of information. When I began watching the videos in extremely slow-motion playback, I saw behaviors I had never seen in years of living with dogs. Examined closely, simple play frolicking between two dogs became a dizzying series of synchronous behaviors, active role swapping, variations on communicative displays, flexible adaptation to others’ attention, and rapid movement between highly diverse play acts.

What I was seeing were snapshots of the minds of the dogs, visible in the ways they communicated with each other and tried to communicate with the people around them—and, too, in the way they interpreted other dogs’ and people’s actions.

I never saw Pumpernickel—or any dog—the same way again. Far from being a killjoy on the delights of interacting with her, though, the spectacles of science gave me a rich new way to look at what she was doing: a new way to understand life as a dog.

Since those first hours of viewing, I have studied dogs at play: playing with other dogs and playing with people. At the time I was unwittingly part of a sea change taking place in science’s attitude toward studying dogs. The transformation is not yet complete, but the landscape of dog research is already remarkably different than it was twenty years ago. Where once there was an inappreciable number of studies of dog cognition and behavior, there are now conferences on the dog, research groups devoted to studying the dog, experimental and ethological studies on the dog in the United States and abroad, and dog research results sprinkled through scientific journals. The scientists doing this work have seen what I have seen: the dog is a perfect entry into the study of non-human animals. Dogs have lived with human beings for thousands, maybe hundreds of thousands of years. Through the artificial selection of domestication, they have evolved to be sensitive to just those things that importantly make up our cognition, including, critically, attention to others.

In this book I introduce you to the science of the dog. Scientists working in laboratories and in the field, studying working dogs and companion dogs, have gathered an impressive amount of information on the biology of dogs—their sensory abilities, their behavior—and on the psychology of dogs—their cognition. Drawing from the accumulated results of hundreds of research programs, we can begin to create a picture of the dog from the inside—of the skill of his nose, what he hears, how his eyes turn to us, and the brain behind it all. The dog cognition work reviewed includes my own but extends far beyond it to summarize all the results from recent research. For some topics on which there is no reliable information yet on dogs, I incorporate studies on other animals that might help us understand a dog’s life, too. (For those whose appetite for the original research articles is whetted by the accounts herein, full citations appear at the book’s end.)

We do no disservice to dogs by stepping away from the leash and considering them scientifically. Their abilities and point of view merit special attention. And the result is magnificent: far from being distanced by science, we are brought closer to and can marvel at the true nature of the dog. Used rigorously but creatively, the process and results of science can shed new light on discussions that people have daily about what their dog knows, understands, or believes. Through my personal journey, learning to look systemically and scientifically at my own dog’s behavior, I came to have a better understanding of, appreciation of, and relationship with her.

I’ve gotten inside of the dog, and have glimpsed the dog’s point of view. You can do the same. If you have a dog in the room with you, what you see in that great, furry pile of dogness is about to change.

It is the nature of scientific study of non-human animals that a few individual animals who have been thoroughly poked, observed, trained, or dissected come to represent their entire species. Yet with humans we never let one person’s behavior stand for all of our behavior. If one man fails to solve a Rubik’s cube in an hour, we do not extrapolate from that that all men will so fail (unless that man had bested every other man alive). Here our sense of individuality is stronger than our sense of shared biology. When it comes to describing our potential physical and cognitive capacities, we are individuals first, and members of the human race second.

By contrast, with animals the order is reversed. Science considers animals as representatives of their species first, and as individuals second. We are accustomed to seeing a single animal or two kept in a zoo as representative of their species; to zoo management, they are even unwitting “ambassadors” of the species. Our view of the uniformity of species members is well exemplified in our comparison of their intelligence. To test the hypothesis, long popular, that having a bigger brain indicates greater intelligence, the brain volumes of chimpanzees, monkeys, and rats were compared with human brains. Sure enough, the chimp’s brain is smaller than ours, the monkey’s smaller than the chimp’s, the rat’s a mere cerebellum-sized node of the primates’ brains. That much of the story is fairly well known. What is more surprising is that the brains used, for comparative purposes, were the brains of just two or three chimpanzees and monkeys. These couple of animals unlucky enough to lose their heads for science were henceforth considered perfectly representative monkeys and chimps. But we had no idea if they happened to be particularly big-brained monkeys, or abnormally small-brained chimps.

^* Of course, researchers soon found brains bigger than ours: the dolphin’s brain is larger, as are the brains of physically larger creatures such as whales and elephants. The “big brain” myth has long been overturned. Those who are still interested in mapping brain to smarts now look at other, more sophisticated measures: the amount of convolution of the brain; the encephalization quotient, a ratio that includes both brain and body size in the calculation; the quantity of neocortex; or the gross number of neurons and synapses between neurons.

Similarly, if a single animal or small group of animals fails in a psychological experiment, the species is tainted with the brush of failure. Although grouping animals by biological similarity is clearly useful shorthand, there is a strange result: we tend to speak of the species as though all members of the species were identical. We never make this slip with humans. If a dog, given the choice between a pile of twenty biscuits and a pile often biscuits, chooses the latter, the conclusion is often stated with the definite article: “the dog” cannot distinguish between large and small piles—not “a dog” cannot so distinguish.

So when I talk about the dog, I am talking implicitly about those dogs studied to date. The results of many well-performed experiments may eventually allow us to reasonably generalize to all dogs, period. But even then, the variations among individual dogs will be great: your dog may be an unusually good smeller, may never look you in the eye, may love his dog bed and hate to be touched. Not every behavior a dog does should be interpreted as telling, taken as something intrinsic or fantastic; sometimes they just are, just as we are. That said, what I offer herein is the known capacity of the dog; your results may be different.

This is not a dog training book. Still, its contents might lead you to be able to train your dog, inadvertently. This will catch us up to dogs, who have already, without a tome on people, learned how to train us without our realizing it.

The dog training literature and the dog cognition and behavior literatures do not overlap greatly. Dog trainers do use a few basic tenets from psychology and ethology—sometimes to great effect, sometimes to disastrous end. Most training operates on the principle of associative learning. Associations between events are easily learned by all animals, including humans. Associative learning is what is behind “operant” conditioning paradigms, which provide a reward (a treat, attention, a toy, a pat) after the occurrence of a desired behavior (a dog sitting down). Through repeated application, one can shape a new, desired behavior in a dog—be it lying down and rolling over, or, for the ambitious, calmly Jet-Skiing behind a motorboat.

But often the tenets of training clash with the scientific study of the dog. For instance, many trainers use the analogy of dog-as-tame-wolf as informative in how we should see and treat dogs. An analogy can only be as good as its source. In this case, as we will see, scientists know a limited amount about natural wolf behavior—and what we know often contradicts the conventional wisdom used to bolster those analogies.

In addition, training methods are not scientifically tested, despite some trainers’ assertions to the contrary. That is, no training program has been evaluated by comparing the performance of an experimental group that gets training and a control group whose life is identical except for the absence of the training program. People who come to trainers often share two unusual features: their dogs are less “obedient” than the average dog, and the owners are more motivated to change them than the average owner. It is very likely, given this combination of conditions and a few months, that the dog will behave differently after training, almost regardless of what the training is.

Training successes are exciting, but they do not prove that the training method is what led to the success. The success could be indicative of good training. But it could also be a happy accident. It could also be the result of more attention being paid to the dog over the course of the program. It could be the result of the dog’s maturing over the course of the program. It could be the result of that bullying dog down the street moving away. In other words, the success could be the result of dozens of other co-occurring changes in the dog’s life. We cannot distinguish these possibilities without rigorous scientific testing.

Most critically, training is usually tailored to the owner—to change the dog to fit the owner’s conception of the role of the dog, and of what he wants the dog to do. This goal is quite different than our aim: looking to see what the dog actually does, and what he wants from and understands of you.

It is increasingly in vogue to speak not of pet ownership but pet guardianship, or pet companions. Clever writers talk of dogs’ “humans,” turning the ownership arrow back on ourselves. In this book I call dogs’ families owners simply because this term describes the legal relationship we have with dogs: peculiarly, they are still considered property (and property of little compensatory value, besides breeding value, a lesson I hope no reader ever has to learn personally). I will celebrate the day when dogs are not property which we own. Until then, I use the word owner apolitically, for convenience and with no other motive. This motive guides me in my pronoun use, too: unless discussing a female dog, I usually call the dog “him,” as this is our gender-neutral term. The reputedly more neutral “it” is not an option, for anyone who has known a dog.

© 2009 Alexandra Horowitz