Dog Sense

by John Bradshaw

  In fact, across the animal kingdom as a whole, the honest display of emotions is favored only in certain quite specific circumstances—namely, when cooperation is the desired result. Humans are among the most cooperative species alive, and indeed many of these special factors apply to us. As a species we evolved in the context of extended-family groups, and so, according to the theory of kin selection, we should tend to be honest with one another. Also, we are extremely good at recognizing other individuals of our own species and recalling our previous encounters with them. Accordingly, we have highly sophisticated cognitive mechanisms for detecting deception among those familiar to us. In other words, most of us are very good at detecting when someone we know is hiding his or her true feelings.

  It’s worth detouring briefly to look at the evolution of human bodylanguage—to see why the connection between facial expression and (some) emotions should be so transparent in our own species—before going on to speculate on whether the same might apply to dogs.

  The human face is particularly expressive. And our facial expressions for the more primitive emotions—such as fear, joy, and anger—appear to be the same the world over. Human expression is clearly a species-typical, evolved trait. The idea underlying this observation was first proposed by Charles Darwin, who attempted to apply the same principle to dogs (see box titled “Darwin’s Dogs”). Indeed, researchers have recently discovered that the facial muscles used to generate these particular expressions are common to virtually all humans, whereas other facial muscles vary widely between races and between individuals. As a species, we cannot function socially without facial expressions—facial paralysis leads almost inevitably to isolation and depression.

  Our facial expressions are directly connected to our emotions. Just watch someone’s face when she’s talking to a friend on the phone—she will smile, frown, and so on just as though the other person could see her. Indeed, one of the functions of our facial expressions—perhaps even their primary function—is to let people know what we’re feeling while we’re talking to them. Under those circumstances, we usually express the emotions that we think we should be feeling, even if we’re not actually experiencing them. The intention in such instances is to convince the speaker not only that we’re paying attention to him but that we are also on his emotional wavelength. Overall, our unconscious facial expressions tend to reassure those around us that we are trustworthy.

  We also consciously use or suppress many of these same facial expressions to modify the behavior of others to our own benefit. There is a direct connection between Emotion III (feelings) and Emotion II (facial expression), but it is under at least some degree of conscious control; indeed, we use our expressions to manipulate those around us. Certain expressions of emotion, such as blushing when we’re embarrassed, are almost impossible to fake or suppress, but some people can produce an apparently sincere smile at will. (Others can manage only a fake smile, technically referred to as a non-Duchenne smile, in which the mouth smiles but the eyes don’t.) We also try to cover up our emotions when it’s socially advantageous to do so: For example, after winning a prize, many people will attempt to block their spontaneous smile by locking their facial muscles or hiding their faces behind their hands, so as to avoid the appearance of gloating. Most of us are adept at detecting false expressions of emotion in others, even though we may not be able to describe precisely how we have detected an insincerity. Completely masking emotion requires considerable practice, as evidenced by the comparatively few individuals who can achieve a “poker face.” Evolution has evidently given us a highly tuned lie-detection system—again, presumably because the success of hunter-gatherer groups depended upon it.

  Darwin’s Dogs

  In his book The Expression of Emotions in Man and Animals, Darwin used the domestic dog extensively to illustrate his ideas about the interpretation of animals’ postures and expressions. One of his central precepts was the “principle of antithesis”—the idea that opposing emotions induce exactly opposite postures, which thereby convey precise information about the animal’s state of mind and intentions. He contrasted the “attacking” dog that stands tall, leans forward with tail and hackles raised, growls, and bares its teeth with the “friendly, submissive” dog that crouches low to the ground, tail held down. Although Darwin’s “principle of antithesis” is rarely referred to today, his interpretation of the dogs’ emotional states is still considered sound.

  Darwin’s “attacking” dog

  Darwin’s “submissive” dog

  Are dogs equally manipulative? Dogs do sometimes appear to be “lying” to each other, especially when there is some conflict of interest involved—although as far as I know, no one has studied this systematically. My Labrador retriever Bruno loved people but was always a bit wary of other dogs, especially other males. When he encountered a person he didn’t know, he’d wiggle his way up to her, half-crouched, his tail twirling round and round like a demented helicopter. When he saw another male dog, he’d stand as tall as he could, and up would go the hackles on his back. In both instances, Bruno was trying to ensure that the meeting would go the way he wanted it to. In the case of the person, he always wanted to make friends, so he used the wolf-cub greeting. He was trying to make himself look smaller than he really was (an effort that rarely succeeded, considering that he was a rather portly Labrador). Toward another male dog, he did the opposite: He tried to make himself look bigger than he really was. Actually, he was bluffing: If the other dog persisted, he’d quickly change tack—he wasn’t very brave—and back away with his tail tucked down. In other words, he appeared to be trying to mislead the other dog. I don’t mean that he was deliberately and consciously setting out to deceive; it is doubtful that dogs have this degree of intelligence. Nevertheless, when attempting to make an impression on a potential rival, most dogs do try to make themselves look bigger than they really are, hoping to scare the other dog off without risking getting hurt in a tussle. However, for this to work every time, the other dog would have to be stupid enough to be taken in by this rather obvious attempt at browbeating. And that seems unlikely.

  So why is it that both dogs don’t simply signal that they have no intention of fighting? In fact, dogs appear to have no way of negotiating such a climb-down. The first few individuals that adopted the tactic of hackle-raising probably gained an advantage from doing this, because their rivals would have been taken in by it. Once most individuals have adopted the habit of raising their hackles, they will expect their rivals to do the same. Any dog that doesn’t raise its hackles will therefore be perceived as smaller than it really is, increasing the probability that it will be attacked. Thus this piece of behavior became fixed in the repertoire; almost all dogs will display it from time to time, even if they have little or no intention of actually fighting. Bear in mind, however, that although their body-language suggests they are bluffing, we have no evidence that dogs are actually aware of this deception—they are simply doing what evolution and their own experiences have told them will achieve the result they want.

  A more complex version of this process probably gave rise to the “bared-teeth” signal that many dogs use as the next stage after hackles are raised. Scientists hypothesize that this strategy originated because actual biting of another dog has to be preceded by pulling the jowls out of the way, to protect them. The “bared-teeth” signal is useful for the other dog as well, because it gives a fraction of a second’s warning that the first dog is about to bite. Presumably, the baring of teeth was often sufficient to forestall conflict: By raising its top lips well before it bit, a potential attacker could force an inexperienced opponent to recoil without having to submit to the risk of an actual fight. But was this really a sensible bluff? The receiver can now see his opponent’s teeth, well in advance of the actual bite. If they’re broken or missing, then he can be confident that the resulting bite may not be particularly painful. Thus both parties gain an advantage from the signal: The attacker shows that he may
be about to bite, and the target can check how damaging this threat is likely to be if carried out. And so this signal, too, gets fixed in the repertoire. Such signals are especially stable, as far as evolution is concerned, because they contain a kernel of honesty in addition to an element of bluff. The attacker is really ready to bite, and the intended victim can really get an idea of what the bite will feel like.

  Bared teeth—an honest signal of fighting potential

  The fact that evolution favors a certain degree of bluffing when two animals are in conflict accounts for why animals’ emotional states may sometimes be difficult to gauge. But dogs, as highly social animals, evidently have more open communication than many other species do. If early dogs (and wolves) were really in a continual struggle for dominance, evolution should have favored a great deal more dishonest signaling and complete masking of emotion—not a good starting point for domestication. By contrast, cooperation, in dogs as well as in humans, tends to favor transparency. For their ancestor the wolf, sustaining the family unit is essential to survival, so it benefits everyone to know how everyone else is feeling. This principle applied equally well to our own hunter-gatherer ancestors. Hence both Homo sapiens and Canis lupus usually show their emotions openly, although wolves (and dogs) use their whole bodies, not just their faces, to communicate their emotional state. This happy coincidence must have been one of the factors that smoothed the path of domestication, enabling each species to learn to read each other’s minds.

  An even greater degree of emotional transparency may have been selected for during domestication, with humans favoring dogs whose body-language was easy to “read” over those that were more inscrutable (that is, until our penchant for unusual features and “baby-faces” started to drive selection in the opposite direction). By implication, those dog owners who are prepared to take the time to learn the signs will find their pets very easy to read.

  Although observing dogs’ behavior and physiological states can offer clues about dogs’ emotions, the connection between physiology and emotion is sometimes murky. A dog’s body-language and, more particularly, its attempts to communicate provide one strand of information as to what it is feeling at any given moment (Emotion II). A second strand comes from its hormones and the activity in its brain (Emotion I): Is the dog internally stressed, elated, or in a state of anticipation? These physiological changes are invisible to owners and are also not yet well studied by scientists, at least not in the dog. Moreover, what is known indicates that such changes often do not correspond one-to-one with a particular behavior or a single emotional state. For example, stress hormones such as adrenaline and cortisol can rise not only when the dog is in a situation it finds uncomfortable but also when it is approaching a potential mating partner. The hormones are preparing the body for activity, not directly reflecting any one emotion. Likewise, most emotions are not simply associated with individual chemicals in the brain. For example, we know that opioid neurochemicals are connected to emotional states, because of the effects on emotion brought about by the narcotics, such as heroin, that mimic them. However, the latter also reduce the perception of pain (e.g., as when morphine is used as an analgesic), so their effects are not simply emotional in nature. Natural opioids—endorphins—are produced in the mammalian brain during social bonding activities such as play and mutual physical contact; the fact that their levels are especially low when the animal is distressed by social separation suggests links to several different emotional states.

  The complexity of these relationships seems to have arisen as the mammalian emotional repertoire evolved piecemeal from that of ancient reptiles, which have much simpler emotional lives (or even, some scientists argue, none at all). Likewise, each emotion is not located in its own unique part of the mammalian brain. Rather, most emotions appear to arise in parts of the midbrain, which is connected to the spinal cord through the hindbrain and, in mammals, is almost completely encased inside the much larger forebrain, the “thinking” part of the brain. Two structures in the midbrain that are key to the generation of many emotions are the hypothalamus and the amygdalae, but these structures are also engaged in other functions, such as hunger, thirst, the sleep-wake cycle, and learning.

  Despite this complexity, it is clear that emotions have a physical presence in the brain and that they are associated with changes in hormones circulated around the body; in short, they have predictable physical manifestations. Thus a combination of the two aforementioned approaches—the physiological (Emotion I) and the behavioral (Emotion II) can be used to investigate which emotions dogs almost certainly possess and which they almost certainly don’t.

  Emotions can be placed in a rough hierarchy from the most primitive (i.e., those that are thought to have appeared first in the evolution of the vertebrates) to the most complex. Since dogs are mammals like us but have less complex brains than our own, it is logical to conclude that we share the simpler emotions but also that the most complex emotions experienced by humans are likely to be ours alone.

  The most basic emotions—such as hunger, thirst, pain, and sexual desire—are perhaps better described as “feelings” than as “emotions.” They are primarily processed by the most primitive parts of the brain—the brainstem, the midbrain, and the hypothalamus. The hypothalamus also processes information relating to reward and punishment; it is therefore crucial to the way that dogs learn.

  The simplest of the true emotions—fear, anger, anxiety, and happiness—are often referred to as “primes.” These are “instinctive” in that they do not have to be learned: No one has to learn how to be frightened; it just happens. They are also “basic” in the sense that they are generated by the most primitive part of the mammalian brain, the limbic system, which appeared very early in the evolution of the vertebrates, perhaps as far back as 500 million years ago. It is therefore almost inconceivable that dogs should not possess these emotions, although it is difficult to gauge precisely what their subjective experience is like.

  Fear, anger, anxiety, and happiness all evolved as ways of responding to significant threats or opportunities. One way of looking at them is to see them as providing “shortcuts.” For example, an animal doesn’t have to scan its memory for the specific threat it is encountering at a particular time and then devise a response; rather, it is prompted by its emotional reaction (fear) to run away quickly, after which it can determine from a safe distance what the threat actually was. This is not to say that learning doesn’t play a part in categorizing such threats more accurately based on accumulated experiences; nevertheless, the underlying emotion will almost always stay the same from one such experience to another.

  Fear may be the most primitive emotion of them all. As for the other simple emotions, the amygdalae, paired almond-shaped structures buried deep in the center of the brain, play a central role in both forming and retrieving memories of frightening events and also in generating the response. The posterior part of the hypothalamus is another key structure, relaying information to and from the brain and out to other hormoneproducing structures, such as the adrenal glands that produce the fightor-flight hormone adrenaline.

  The expression of fear in dogs follows a pattern that is recognizably similar to the expression of fear in man. It usually begins with the dog becoming suddenly alert and then freezing, rooted to the spot while the amygdalae furiously signal to the cortex, the “thinking” part of the brain, for the correct response to the situation. Meanwhile, the dog holds itself tensely, possibly shaking visibly, with eyes wide and teeth bared, as a general preparation for all things dangerous. Beneath the skin, the heart rate and breathing both speed up.

  If the situation is unprecedented, there may be nothing helpful in the memory databank. This can lead to behavior that may seem downright bizarre to us more logical humans. For example, a dog that has never seen a cardboard box before in its life may fail to identify it as a harmless inanimate object and, by default, go into a full-blown fear response.8 Fear is a shortcut for categor
izing events, and what falls into the “scary” category depends upon what the dog has experienced before, and what it hasn’t, especially during the first six months of its life. The “scary” category will consist of two sorts of things: those that have frightened the dog in the past, and those the dog has had no experience of whatsoever.

  The dog’s response to the scary situation will depend on what it has found to work best in the past. Some dogs will almost always freeze; others will usually run away. Still others, especially if their escape route has been blocked on previous occasions, may resort to aggression almost immediately. Indeed, many clinicians will tell you that most of the cases of aggression they see are motivated by fear—not by anger, or any need to “dominate.” Fear also lies at the heart of many other behavioral disorders. It is arguably the most powerful of the emotions that dogs possess.

  Fear is a powerful trigger of learning. Dogs that are suddenly frightened by something unfamiliar, such as a cardboard box, are likely not only to continue to be frightened by similar boxes but may also show palpable signs of apprehension when they revisit the place where the original fright occurred, even though the scary object is no longer there. This is one way that dogs can develop what appear to us to be “irrational” fears—although they presumably make perfect sense to the dog, who is recalling the whole event, not just the “obvious” unfamiliar stimulus.

  Anxiety is sometimes confused with fear, in that it shares some of the same manifestations. But anxiety is about the anticipation of fear—it is triggered not by an actual object or event that is intrinsically frightening but, rather, by predictors of a frightening event that may occur at some indeterminate time in the future. My first two dogs, Alexis and Ivan, both Labrador/terrier crosses (Jack Russell and Airedale respectively), were self-confident to the extent that I doubt they ever felt much anxiety. My third, Bruno, was a purebred Labrador, and an altogether more emotionally dependent animal—not easily frightened but very reliant on the humans around him. Before he arrived at our house as an eight-week old puppy nearly thirty years ago, I had never heard of “separation anxiety”—nor had many veterinarians, ours included. Fifteen years later, I started a research program that revealed, among other things, that half the young Labradors in the UK hate being left alone; but in those days I don’t think anyone even suspected that this was the case.