In Defence of Dogs

by John Bradshaw

  We also consciously use or suppress many of these same facial expressions to modify the behaviour 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 are embarrassed, are almost impossible to fake or suppress, but some people can produce an apparently sincere smile at will. (Others can only manage a fake smile, technically referred to as a non-Duchenne smile, in which the mouth smiles but the eyes do not.) We also try to cover up our emotions when it is socially advantageous to do so: for example, after winning some 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. Masking emotion totally requires considerable practice, as evidence 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.

  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 people he did not know, he would wiggle his way up to them, half-crouched, his tail twirling round and round like a demented helicopter. When he saw another male dog, he would 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 (which rarely succeeded, considering that he was a rather portly Labrador). Towards 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 would quickly change tack – he was not very brave – and back away with his tail tucked down. In other words, he appeared to be trying to mislead the other dog. I do not mean that he was deliberately and consciously setting out to deceive; it is doubtful that dogs have this degree of intelligence. Nevertheless, 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 do not 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 does not raise its hackles will therefore be perceived as smaller than it really is, increasing the probability that it will be attacked. So this piece of behaviour 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 that 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.

  DARWIN’S DOGS

  In his book The Expression of the Emotions in Man and Animals, Charles Darwin used the domestic dog extensively to illustrate his ideas on the interpretation of animals’ postures and expressions. One of his central precepts was the ‘principle of antithesis’, the idea that opposing emotions induced exactly opposite postures, which thereby conveyed 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, with 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

  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 lip 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 are 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 favours 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, then evolution should have favoured a great deal more dishonest signalling and a complete masking of emotion – not a good starting point for domestication. By contrast, co-operation, in dogs as well as in humans, tends to favour 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 alike 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 the other’s minds.

  An even greater degree of emotional transparency may have been selected for during domestication, with humans favouring 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’ behaviour 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 invisib
le 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 behaviour 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 (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). As a consequence, each emotion is not located in its own unique part of the mammalian brain. Rather, most emotions appear to arise in various 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 do 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 or Emotion I, and the behavioural or Emotion II – can be used to investigate which emotions dogs almost certainly possess, and which they almost certainly do not.

  Emotions can be placed in a rough hierarchy, from the most primitive (that is, 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 parts 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 ‘short-cuts’. For example, an animal does not have to scan its memory for the specific threat it is encountering 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, of course, that learning does not play a part in categorizing such threats more accurately, based on accumulated experience; 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 centre 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 hormone-producing structures, such as the adrenal glands that produce the fight-or-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 bank. This can lead to behaviour that may seem downright bizarre to us more logical humans. For example, a dog that has never seen a large 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 short-cut for categorizing events, and what falls into the ‘scary’ category depends upon what the dog has experienced before, and what it has not, 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 that they see are motivated by fear, not by anger or any need to ‘dominate’. Fear also lies at the heart of many other behavioural disorders, and is arguably the most powerful of the emotions that dogs possess.

  Fear is also a powerful trigger of learning. Dogs that are suddenly frightened by something unfamiliar, such as a cardboard box, are not only likely 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, which 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. When 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 veterinary surgeons, ours included. Fifteen years later, I started a research programme that revealed, among other things, that half the young Labradors in the UK hate being left alone, but in those days I do not think anyone even suspected that this was t
he case.

  Bruno could not hide the anxiety he felt whenever he realized that we were about to go out. Locating of car keys, putting on coats, collecting children from the four corners of the house – these actions triggered an expression of absolute misery on his face, and he slunk off to his bed, the place he felt most secure. The ‘experts’ at the time told us that this was only a game he was playing to stop us going out, that gundogs were bred to be left in kennels for hours at a time and were perfectly happy doing so. As soon as we were gone, we were told, he would settle down and sleep until we returned. Wrong: his ongoing anxiety was obvious from the chewed-up bed, furniture, even wallpaper, that we found when we returned home.

  These are all signs of anxiety. Retrievers are very mouth-focused, and chewing seems to be their favourite way of relieving tension; if Bruno had been a different type of dog, he might have barked, paced, scratched at the walls, or urinated or defecated on the floor. When we tried to put him in boarding kennels, where the opportunities for chewing were limited, he turned to howling – for hours at a time. In the end we accepted that he was just a very attached dog, and tried to make sure that he was always with someone he knew. He was not frightened of our going out, but he knew that he hated being left alone, so the signs that told him that he was about to be left made him anxious. He was probably also anxious that we would never return; dogs’ concept of time is not fully understood, but seems less precise than our own, so it is difficult to know how much they can anticipate things that might or might not happen at some indeterminate time in the future.