How Dogs Think: Understanding the Canine Mind

CHAPTER 1

The Mind of a Dog

I myself have known some profoundly thoughtful dogs.

—JAMES THURBER

PALEONTOLOGISTS TELL US that 14,000 years ago a Stone Age man sat next to a fire looking at an animal that we would readily recognize as a dog if we were to see it now. This forefather of all of today’s dogs was not just a household pet. He was a sentinel, protector, and hunting partner. His descendants would become, among other things, shepherds, comrades in war, search-and-rescue heroes, law enforcement officers, lifeguards, guides for the blind and deaf, assistants for the disabled, as well as valued family members and companions. One can imagine that this early man, who had just learned to make weapons and tools out of sticks, stones, and bits of bone, might have paused to look into the dark, soulful eyes of his companion and wondered, What is he thinking about? How much does he know? Does he really have feelings, and if so, what does he think about me?

It is now 140 centuries since that fire last flickered and died, and we are still asking those same questions about our canine companions. The average dog owner still looks into the eyes of his dog, sees what appear to be sparks of intelligence, emotion, and awareness, and wonders about what is going on in the dog’s mind. Historically, many great thinkers have wrestled with this question. Some have intuited that there is a knowing awareness behind those eyes, while others have concluded that dogs merely act according to genetically programmed instincts.

Dogs and Philosophers

The Greek philosopher Plato had a very high opinion of the dog’s intellect. He described the noble dog as a lover of learning and a beast worthy of wonder. In one of his dialogues he presents a discussion between Socrates and Glaucon in which Socrates, after much analysis, eventually convinces his disciple that his dog is a true philosopher.

Plato’s contemporary Diogenes, another significant Greek philosopher, although more eccentric than most, became known for wandering the world with a lamp claiming to be looking for an honest man. While he had his doubts about humans, Diogenes thought dogs were extremely moral and intelligent and even adopted the nickname Cyon, which means Dog. He would go on to found one of the great ancient schools of philosophy, and he and his followers would become known by his nickname as Cynics or Dog Thinkers. Diogenes’ own intelligence and wit were such that Alexander the Great, after meeting him in Corinth, went away saying, If I were not Alexander, I should wish to be Diogenes.

When Diogenes died, the people of Athens raised a great marble pillar in his memory. On top of the pillar was the image of a dog. Beneath the dog there was a long inscription that started with the following bit of conversation:

Say, Dog, I pray, what guard you in that tomb?

A dog.

His name?

Diogenes.

There are many times when the behavior of my own dogs brings me back to the admiring views of Plato and Diogenes. One cold rainy day, when I was feeling too tired and uncomfortable to take my dogs on their usual morning walk, they had to content themselves with being let out in the yard for a short while. For my flat-coated retriever, Odin, this simply was not an acceptable situation and, late in the afternoon, I was disturbed from my reading by a clatter at my feet. I looked down and noticed that Odin had somehow found his leash and deposited it on the floor. I picked it up, put it on the sofa next to me, and gave him a pat and a reassuring Later, Odin.

A few minutes passed and there was another clatter at my feet; I found that Odin had now deposited one of my shoes beside me. When I didn’t respond, he quickly retrieved the other shoe and put it down next to me. Obviously, to his mind, I was being quite dense or stubborn, since I still delayed going out into the cold and wet weather. It was at that moment that Odin ran to the door and gave a familiar bark. It was a distinctive sound that he only used when my wife, Joan, was approaching the door. I had spent several years teaching at a university in New York City and had developed the habit typical of New Yorkers, which involves always locking doors, even on days when I was inside working at home. This tended to annoy Joan, who grew up in the safer and less paranoid environment of Alberta, Canada. So when Odin gave his Joan is here bark, I got up to unlock the door rather than leave her fumbling for her keys in the rain and getting annoyed with my inconvenient habit. The moment I got within a foot or two of the door, Odin dashed back to the sofa and grabbed his leash. Before I had even determined that Joan’s car had not arrived in its usual place, he was nudging my hand with the leash he carried in his mouth.

I started to laugh at his subterfuge. I could imagine his mental discourse of the past few minutes running something like I want a walk, so here’s my leash.—OK, I’ve brought you your shoes, so let’s walk.—All right now, while you’re already standing at the door, and while I’m now offering you the leash, why don’t we just take that walk? I have obviously added to Odin’s behavior a whole lot of reasoning, an internal dialogue, and the idea that there was some kind of conscious planning involved; however, these behaviors certainly would have been consistent with his actions. And by the way, he did get his walk.

Minds More or Less

Although the idea of an intelligent, reasoning, and feeling dog persisted for many centuries, we might say that in the seventeenth century dogs lost their minds. However, according to one of the most influential French philosophers of the time, René Descartes, dogs had no minds to lose. An exceptional mathematician, Descartes also performed some important experiments in physiology, but it seems likely his strong Catholic religious feelings, not his scientific findings, led him to this conclusion. To Descartes, granting dogs any degree of intelligence was equivalent to admitting that dogs had consciousness, which would include awareness and the ability to plan future actions. According to religious doctrines at the time, however, anything that had consciousness also had a soul, and anything that had a soul could earn admission to heaven. That dogs might go to heaven was unacceptable to both Descartes and the Roman Catholic Church at that time.

This left Descartes with the problem of explaining how, in the absence of intelligence, reasoning, or consciousness, dogs could have such complex behaviors. His answer came when he visited the gardens of Saint-Germain-en-Laye, the birthplace and home of Louis XIV. Those gardens featured the seventeenth-century equivalent of animatronics or robotics in the form of elegant statues designed by the Italian engineer Thomas Francini. Each figure was a clever piece of machinery powered by hydraulics and carefully geared to perform a complex sequence of actions. Thus one statue might play a harp, while another danced, and so forth. Descartes reasoned that dogs might be the biological equivalent of these animated machines, but instead of being driven by hydraulics and gears, they are controlled by physical reflexes and unthinking responses to things that stimulate them. The observation that dogs respond to their environment does not invalidate his argument, since those statues also responded to outside events, such as when a person stepped on a particular paving stone, which triggered the switch that in turn activated the statue.

Although Descartes’s view of dogs as unreasoning and unconscious bits of biological machinery dominated scholarly thinking for two centuries, it received a major challenge in the middle of the nineteenth century, when Charles Darwin and his theory of evolution emerged upon the scene. Dissenting from the common teaching that each species had been separately created by God, Darwin concluded that humans were not special or unique in their mental abilities. In his book The Descent of Man, he stated that the only difference between man and most of his lower mammalian cousins is one of degree and not of kind. He went on to say that the senses and intuitions, the various emotions and faculties, such as love, memory, attention, curiosity, imitation, reason, etc. of which man boasts, may be found in an incipient or even sometimes in a well-developed condition, in the lower animals.

Darwin described animals and people as part of a continuum in evolution that gave rise to different levels of awareness, reasoning ability, intelligence, and memory in different species. Since, according to him, these are the components of consciousness, different species would also have different levels of consciousness. Thus a dog might be conscious and self aware, but not to the same degree as a human being.

Recent research has supported Darwin’s view by demonstrating similarities between the nervous systems of dogs and humans. For instance, researchers have proven that the nerve cells in a dog’s brain work the same way as those in a human brain. The neurons that make up the human brain have the same chemical composition as the neurons in a dog’s brain, and the patterns of electrical activity are identical. The structure of a dog’s brain contains most of the same organs that are found in the human brain.

Like humans, dogs have special areas of the brain that control specific activities. In fact, if we drew a map of the locations of various functions in a dog’s brain, it would be remarkably similar to the map of those same functions in the human brain. For instance, in both dogs and people vision is located at the very back of the brain and hearing is located at the sides of the brain, near the temples. The sense of touch and control of movements are located in a thin strip running over the top of the brain in both dogs and humans.

Even more striking is data from a recent scientific study from the dog-genome project. Ewen Kirkness of the Institute for Genomic Research in Rockville, Maryland, and his research team compared the DNA of a poodle to that of a human. What they found was that there was more than a 75 percent overlap between the genetic codes of humans and canines. All of this physiological similarity is certainly consistent with the belief that there should be a lot of similarity in the behaviors and workings of the minds of humans and dogs.

Minds Lost and Found

In the science of animal minds or animal consciousness, the pendulum tends to swing from one extreme to another. Although Darwin’s views never completely disappeared, in the early part of the twentieth century a new psychological perspective called Behaviorism held sway. It was an idea more compatible with Descartes than Darwin. This new way of looking at animal behavior was the brainchild of psychologist John B. Watson, at Johns Hopkins University, and would be perpetuated into the present by the research of the well-known psychologist B. F. Skinner at Harvard University.

Behaviorists believe that the only aspects of behavior that can be legitimately studied are those that can be observed and measured by an outside third party. For them, even to talk about consciousness or mind, especially when considering the behavior of animals, is only empty speculation, since we can’t really measure awareness, feeling, or thought. Behaviorists believe that we don’t even need ideas of consciousness or thought to explain behaviors. No doubt Dr. Watson would have vigorously objected to the mentalistic interpretation of Odin’s behavior that I gave earlier.

To illustrate the behavioristic view: Suppose that I told Watson that my dog likes meat. He would likely say that this is a projection of my own feelings. He would note that if I gave the dog a piece of meat, it is legitimate to describe the dog’s behavior—the fact that he barks, jumps up, wags his tail, drools, opens his mouth, and eats the meat—but it is not legitimate for me to say that the dog likes the meat, or wants the meat, or is aware that the meat will soon be given to him. All such conclusions are merely projections of attitudes or feelings by the human observer onto the dog, and Watson would contend that there is no evidence to support statements that my dog has any thoughts or feelings at all. Like Descartes, the scientist would simply observe that Odin’s activity level rises when he is given meat and he performs certain behaviors in response to getting it.

As the field of animal cognition expands, modern psychological theory is becoming a bit more accepting of animal minds again. Biologists and psychologists are starting to talk about purposive behaviors and to openly discuss the possibility that dogs might have a true conscious representation of their world. The battle between the dog as thinker and dog as machine continues, however, and can be heard in the hallways and laboratories of many behavioral science departments around the world. Sometimes these arguments can get quite heated, and often they appear to take on some of the qualities of arguments about religion—the issue becomes one of belief and emotion, rather than of scientific fact.

Canine Mental Abilities

Regardless of whether people grant dogs a mind with consciousness, reasoning, and complex thought, or whether they insist that dogs are simply machines that use neurons rather than silicon chips to process information, there are certain things that can be agreed upon as scientific facts:

Dogs sense the world and take in information from it.

Dogs learn and modify their behavior to fit circumstances.

Dogs have memories and can solve certain problems.

Early experiences as a puppy can shape the behaviors of the adult.

Dogs have emotions.

Individual dogs seem to have distinct personalities and different breeds seem to have different temperaments.

Social interactions, including play, are very important to dogs.

Dogs communicate with each other and with humans.

Unfortunately, each of these agreed-upon facts raises a number of new questions. For instance, do dogs perceive the world the same way that we do? If not, how does the world appear to them? What can we sense that dogs can’t and what can they sense that we can’t? Are dogs’ memories different from ours? What kinds of problems can dogs solve, and at what point does their intellect start to fail them? Can dogs understand time, beauty, music, or arithmetic? Do dogs really have ESP, as some people have claimed? When we talk about the temperament of dogs, are we talking about the same thing that we call personality in people? Can dogs learn to do things by simply observing others doing them? All these questions, and myriad others, can be answered even if we can’t yet determine whether dogs have true thought and consciousness or whether they are merely fur-covered computers.

Alien Minds

Most people reading this book have much the same hope that I have, namely that we can learn what our dogs are aware of and what and how they are thinking. Realistically, however, we must recognize that this goal may prove to be unattainable and we may never understand the mind of a dog as completely as that of another human being.

The main problem is that we are human and as such we can reason only as a human does. If the experience of the animal is completely different and alien to ours, we may have no human reference that allows us to interpret the thinking behind the behavior. Imagine trying to derive how the world appears to a bat, flying through the night, guided only by its natural form of sonar, or how the world appears to a nematode, a tiny worm with no hearing or vision but only a chemical sense and a primitive sense of touch. In each case we might fall back on familiar human experiences, such as supposing that the bat’s consciousness during sonar navigation must be like a person’s with his eyes closed trying to locate objects in the world by sounds and echoes, but we would be speculating. The bat’s sonar might give him a full, rich experience of a world full of objects much like the high-tech sonar used on modern ships, which uses sound reflection and high levels of computer analysis to provide a detailed map of the ocean floor. How that might translate into consciousness is difficult to imagine. What about the experience of the nematode? Can you imagine tasting your way through the environment? Can you build a map of the world in your mind by simply tasting the chemical concentrations that you pass through? We simply may not have the mental capacity to imagine what a nonhuman mind senses and thinks.

Can Mental Process Be Observed?

Even the most careful scientist will have a natural tendency to interpret all behaviors in human terms. Unfortunately, the cause or reason behind a particular human behavior may not be the same reason behind an apparently similar animal behavior. Consider the following simple situation: We tell a child that we have a problem that she has to solve, but we don’t give her any details. Next we give her two cards out of a group lying face up on the table. In this case, each of the cards has the number 2 written on it. Now we ask her to find an answer. The child responds by going over to the remaining cards and picking out one that has the number 4 on it. Next we bring a dog into the testing room and present the same two cards to him. Much to our surprise, he responds by also selecting a card with the number 4 on it. Since we can’t directly interview the dog, we consult the human child to determine what mental processes were involved. When we ask the child How did you get the answer 4? she replies that she took the first number and added to the second number and the sum came out to be 4.

On the basis of our knowledge about how a human solves the problem, can we conclude that the dog that we are testing knows how to add? While that might be tempting, from a scientific viewpoint it is a grave error. To begin with, we view the figures 2 and 4 as numbers, but a dog views them as nothing more than abstract patterns. It is also possible that although we think that the most relevant thing about the cards is the number written on them, there might be other things associated with the cards or the testing situation that the dog responds to. Perhaps the dog’s behavior is a lot more subtle than just glancing at the figures written on the cards. For instance, the dog could be looking more intently at us than at the cards. Because we expect (or hope) that the dog has arithmetical skills, and since we are adding these numbers in our heads, we might glance directly at the card that has the 4 on it. The dog, with no knowledge of arithmetic, simply follows the direction of our gaze and then goes to the card that we looked at.

Since the dog views the world quite differently than we humans do, and has different sensory capacities and different priorities, our canine test subject may not actually be working on the mental problems that we have set for him. The dog may well be using mental processes and sources of information that simply don’t occur to us. For example, when my children were quite young I had a cairn terrier named Flint. I convinced my children that Flint could read by giving them the following demonstration. First I asked them to draw something. Let’s say they drew a cat. Next I had them print three words on three separate pieces of paper; let’s say they wrote cat, dog, and horse. I then folded each piece of paper into a sort of a tent, shaped like a ©, with the word facing toward the dog. Next I showed the dog the piece of paper with the picture of the cat and in serious tones I explained, Flint, this is a cat. Go find the word that says ‘cat.’ In response to this, Flint would dash off with a little yip of delight and would always bring the piece of paper with the correct word written on it back to me. We would repeat this several times, with different pictures and different words. On the basis of this test, my children became convinced that Flint could read. I even managed to convince some psychologist colleagues of mine that somehow I had taught the dog to associate the shapes of the letters in simple words with either the appropriate image or the sound of my voice saying the word.

The truth of the matter was that I was cheating, using the same kind of misdirection that stage magicians use. Before I would let Flint perform his reading demonstration, I would stop in the bathroom and scrape the nails of my left hand across a bar of soap. When the children drew the picture, I transferred it to my left hand and put a bit of soap on it. I did the same thing with the paper containing the correct word while holding it in my left hand. My right hand, with no soap on it, was used to carry the papers with the wrong words on it. When I was describing the picture to Flint and saying the word, I held the paper up near his nose, so that he could smell the soap. When he dashed out to read the correct word, he was simply seeking another piece of paper that smelled exactly like the one that I placed before him. That solution never occurred either to my children or to my amazed professional colleagues. They viewed the situation as humans, whose dominant sense is vision. The fact that such a minor human sense, such as smell, could really hold the answer did not occur to them.

Thus, returning to our earlier example with the dog and the two cards with the number 2 on them, it could be that that dog had solved the problem in a manner that no human could. Perhaps, like Flint, the dog was solving the problem using scent. Since the child had handled both of the cards carrying the number 2 and also handled the card with the number 4, these cards now carry that child’s scent. Thus the dog may be reading the problem that we are setting for him as asking him to find another thing that smells just like these two. Or perhaps he is using some other mode of thinking that we have no way of fathoming.

This is the caution that we must hold in our minds. The same behavior can result from completely different processes. With human subjects, if we are asked to predict a boy’s reaction when a girl that he fancies tells him No, we must first know what question he was asking! Similarly, when we try to interpret what a dog is thinking or doing, we must know how he interprets the question, what he is trying to accomplish, and which methods and processes he is using. There is no guarantee that his mind will operate like a human mind when faced with a similar situation.

As a first step in describing the mental processes of a dog, we must understand what his world looks like to him and this requires a knowledge of his senses. Some predatory fish hunt their prey by sensing the vibrations or currents stirred up by animals moving in the water near them. Honeybees are guided to their source of nectar because they can see in the ultraviolet range of light (which is beyond human visual ability). Many flowers mark the source of nectar and pollen in patterns that look like dark bull’s-eye targets when viewed by an eye that sees ultraviolet light. Many hunting snakes can see in the infrared range (which people cannot). The body heat of warm-blooded animals is a form of infrared energy, so on the darkest night they appear to be bright glowing beacons to the snake. The worlds of these different animals could appear very different from the world that we humans see, since they use senses that either we don’t have or we use in a different way.

Since there are significant differences between a dog’s senses and a person’s, we must first explore what passes through the dog’s senses. It is this sensory information that determines the canine view of reality and ultimately shapes the dog’s way of thinking about the world.

CHAPTER 2

Getting Information into the Mind

SINCE THE FIRST experimental psychology laboratory was started in Leipzig, Germany, by Wilhelm Wundt in 1873, psychologists have tried to understand the components of human consciousness by performing experiments on the senses of vision, hearing, and touch. Without these senses and those of taste and smell, your brain, which is responsible for your conscious experience, would be an eternal prisoner in the solitary confinement of your skull. Data from the senses are the building blocks upon which thoughts and minds are constructed. The Greek philosopher Protagoras summed up this notion around 450 B.C. when he said, We are nothing but a bundle of sensations.

If this notion bothers you, then consider the mind as if it were a computer. Mental operations are the way in which the computer processes data. Obviously, the data the computer processes must come from somewhere. In animals, information comes into their mental computers from the senses, while the data for the computer might come from the keyboard. Now imagine for a minute that a certain kind of computer has a keyboard with only numbers on it. No matter how powerful that computer is, the only data that it can ever process will involve numbers. A keyboard that contains letters as well as numbers changes the nature of the data that a computer can receive, enabling it to process words and language. But if there is no letter B on the keyboard, then the words brain and rain are not distinguishable. This means that we cannot identify whether the word received by the computer refers to a part of the body or a weather condition. Similarly, if there are limitations on the sensory abilities of an animal, this will affect or bias the conclusions that the animal’s mind is capable of reaching. Our senses set limits on the kind of data that our minds can process; thus instead of lacking a B, an animal may lack the ability to discriminate among colors, which would prevent him from learning things about the world if the most useful information is based upon color differences. If some senses are stronger than others, this may bias us toward seeking out certain sources of information where the sensory data are better and might cause us to ignore sources of information where the data are not very good. Thus we humans have the saying Seeing is believing (because vision is our most precise sense), while for dogs it might be Smelling is believing.

Movies in the Mind?

Most people remember very little that happened before the age of four years. Certainly you were learning many things during the first three or four years of your life, and you did have many experiences that should have resulted in memories. You were probably toilet trained then; you learned how to eat with utensils and how to recognize your parents and other family members; and you probably knew a few children’s games and had been to a few interesting places, like the zoo or the seashore. You did form memories of those events, but they were coded or registered in your brain as visual images. Once you developed more advanced language abilities, you changed the way you recorded memories and began to register nearly all of them in the form of a language code. You also began to think by using language rather than images. Those early memories are still there, but because you now think in words, you can’t retrieve them—you have lost the key to that part of your past that was registered in images rather than in words.

People who have learned more than one language often find that it is possible to think easily about certain topics only in the language that they were using when they first learned that information. I have a friend who is bilingual in French and English but who was courted by her husband when she met him during her schooling in Paris. Although they now live in an English-speaking culture, they still find that they can talk about their intimate feelings only in French, the language they were speaking when they fell in love! If I say personal things to Pascal in French, they have feeling and meaning, she told me, but if I say them in English, they seem fake—maybe even funny. On the other hand, when we talk about things that have happened to us since we began living in Vancouver, we always speak in English. It is as though I can’t find the French words for these things.

Dogs do not have language, or at least not the word-based language that humans use. This means that their thoughts will be coded in a form that is quite different from that in humans. In the absence of language, dogs must resort to mental processes that may be similar to the sensory-based thinking that humans use as toddlers. I am not suggesting that dogs are running videos inside their heads, but rather that the substitute for words in their thought processes is a set of images drawn from the experiences that their senses provide them.

Some scientific data suggest that when dogs think about something, their brains act in a manner very similar to the sensory experience of that situation. V. S. Rusinov, a Russian scientist who used dogs to study the electrophysiology of the brain, mounted some sophisticated equipment that measured the brain waves of dogs and transmitted them, via tiny transmitters, to recording devices. Each day, the dogs were brought into the laboratory for various training and perceptual experiments. This was done on a schedule, so for five days each week the dogs began their testing session at the same time. Rusinov soon noticed that as the dogs first came into the testing room, their brains showed electrical patterns that indicated they were fairly relaxed. Once the laboratory session started, however, there were some characteristic changes in the brain wave pattern that seemed to be associated with what the dogs were experiencing. One weekend Rusinov brought a group of visitors into the lab and, although he had no intention of actually testing any dogs that day, he turned on the equipment that normally recorded the brain wave patterns, just to show the visitors how it worked. Much to his surprise, the dog that was normally scheduled for tests at that time on weekdays was producing brain wave patterns that were nearly identical to his regular working patterns. Once the testing time was past, however, the dog’s brain waves returned to their normal, nonworking patterns. Rusinov’s conclusion was that the dog was sensitive to the passage of time, and when the time came for his testing, he began to think about the testing situation and what normally happened then. Since dogs register memories as sensory images, their thinking activates some of the same regions of the brain that would be used if the dog were actually seeing and hearing what he usually experienced in the lab.

Obviously, if the stuff that a dog’s thoughts are made of consists of particular sensory ideas, and sense images take the place of words in the dog’s thinking, then we can understand the dog’s thought processes only if we understand the language of his senses. If he fails to sense things as humans do, then he will fail to reach the same conclusions that people do when presented with a particular situation. If he senses things differently from the way humans do, or senses things that humans do not perceive, then his thoughts and understanding of what is going on will be quite different from ours. Thus if we want to learn the language of the dog’s mind, we must learn the vocabulary provided by his senses.

The Eyes of a Night Hunter

Human and canine eyes are built around the same general design, but they have significant differences that affect how each species perceives the world. In humans the visual system uses a greater portion of the brain and more neurons to process and transmit information than do any other senses, which means that our interpretation of a situation is usually biased toward what we can see. This is not the case for dogs. Their visual system does not dominate their brains to the same degree, and their interpretation of the environment is less strongly dependent upon sight.

Although in many ways the dog’s visual abilities are poorer than people’s, the dog does have relatively good vision by animal standards. In some ways the dog’s visual processing ability is actually better than that of a human. Certainly, nobody who has ever seen a dog follow the flight path of a thrown Frisbee, and leap into the air to catch it, can doubt that the dog is processing images and has good visual ability. In many dog breeds the ability to do the jobs for which they have been bred and trained depends on their vision. A retriever needs to track birds visually and to mentally mark the places where they fell after a hunter shoots them. Sheepherding dogs can detect small movements of members of a herd, which inform him where he must move in order to keep the flock together. He also needs to be able to see from a distance the hand or arm signals that his master makes to indicate which direction he wants him to go and where to move his flock. Sight hounds, such as the greyhound, have been bred for thousands of years to pursue and catch swiftly running game based solely on vision. Guide dogs for the visually impaired use the visual data they receive through their eyes to substitute for the visual information that their masters cannot detect.

Evolution has fine tuned the senses of every species so that the species can survive. Because humans evolved from tree-dwelling primates, we needed eyes that could see colors (to pick out ripe fruit and nuts from among the leaves of trees), good visual acuity (to see small nuts and berries), and good depth perception (so that we would not misjudge the distance between branches and fall to the ground). The ancestors of dogs were primarily hunters and meat eaters that were adapted to run swiftly on the ground to pursue prey that might be distant but still within chasing range. Canines are also crepuscular, meaning they are usually active at dusk and dawn and are more comfortable than humans when operating in dim light. The type of eye needed for twilight and nighttime activity requires sensitivity to low levels of brightness, but perception of color is really not very important.

Understanding a dog’s visual sense requires some understanding of the anatomy and physiology of the eye, which operates much like a camera. Both the eye and a camera require a hole to let light in (the shutter aperture in the camera and the pupil in the eye), a lens to gather and focus the light, and some kind of sensitive surface to register the image (the film in the camera and the rear surface or retina in the eye). Both the eye and the camera need features to allow them to adjust to various light conditions, and both are continually making compromises between achieving the maximum ability to pick up low levels of light and achieve the maximum ability to see small details.

At every stage in the construction of the dog’s eye, however, the choice seems to have been made to sacrifice a certain amount of detail-resolving ability in order to function better at low light levels. For instance, a dog’s pupils are much larger than those in most humans. In many dogs you can’t really see much of anything except the wide pupil filling the eye, with just a hint of colored iris around the edge. While this lets in more light, such a large pupil also results in a loss of depth of field, which is the range, or near-to-far distance, over which objects are in clear focus. To return to the camera analogy, if a photographer wants to blur out the background behind someone he is photographing, he will use a wide aperture (f-stop) to reduce the depth of field. If he wants everything in focus, including the mountains on the distant horizon, he will use the smallest aperture possible. A dog’s pupils can enlarge or contract too, but dogs cannot make their pupils small enough to give them the same depth of field that humans have.

The dog has more light-gathering power in his eyes because of larger lenses. To gather a lot of light, a lens has to be big, which is why astronomical telescopes, such as that at Mount Palomar in California, can have lenses as large as 200 inches (500 cm) across. There are effectively two parts of the eye that serve as lenses in humans and dogs. The first is the cornea, the transparent portion of the eye that bulges out at the front, which is responsible for the actual light gathering. The second, the crystalline lens, is behind the pupil and is responsible for changing the focus of the light. Animals that are active in dim light usually have large corneas. Notice how large your dog’s corneas are in comparison to those of humans.

The light that passes through the pupil and the crystalline lens eventually form an image on the retina. Here much of the light is caught and registered by special neural cells called photoreceptors. As in human beings, there are two types of photoreceptors in the retina: rods, which are long and slim, and cones, which are short, fat, and tapered. The rods are specialized to work under dim light conditions. Not surprisingly, dogs have a much higher proportion of rods in their eyes than humans do, but they also have an additional mechanism to meet the needs of night hunting that is not found in humans.

You might have noticed that at night, when a dog’s eyes are