In Defence of Dogs

by John Bradshaw

  We can be reasonably sure that there was a deliberate element in the breeding of all these dogs, over at least the last 5,000 years, by the simple expedient of allowing bitches to mate only with chosen males of similar type. Some males were evidently favoured over others: molecular biologists have found much more variety in the (maternal) mitochondrial DNA of dogs than in the Y-chromosome (paternal) DNA, indicating that, during the entire history of the dog, far fewer males than females have left surviving offspring. Favoured males must therefore have been prized and taken to mate with many bitches, much as happens today within pedigree breeds. The choice of male must sometimes have been based on body conformation (for example in dogs bred for food) but mainly it would have been based on whatever kind of behaviour was desired in the puppies, whether that be suitability for herding, hunting or guarding.

  Native American dog travois

  Although dogs were almost certainly being bred deliberately 5,000 years ago, matings based on the dogs’ own preferences would have kept the dog population diverse. Since dog-keeping would have been much more chaotic than it is today, many matings would also have been unplanned – and if the resulting offspring turned out to be useful, they would have been retained. Taboos against raising puppies that were not ‘purebred’ would have been rare, unlike the situation today. So without any deliberate planning, a healthy level of genetic variation was maintained within types, as well as between. Transfer of dogs from one location to another by traders would have ensured that most local populations were not reproductively or genetically isolated from one another, maintaining diversity at the local as well as global levels. In the absence of veterinary knowledge, natural selection would have continued as a major force directing the development of dogs in general; the rates of both reproduction and mortality would both have been much higher than they are today, at least in the West. Dogs that were prone to disease or infirmity, or carried other disadvantages, such as difficulty in whelping, would have left few offspring, and their lineages would eventually have died out.

  As the modern world developed, so did the degree of deliberate breeding, for purposes that were at the same time both increasingly diverse and narrow in definition. For example, further specialization within the existing range of sizes and shapes occurred in medieval Europe, where the importance to the new aristocracy of hunting led to the breeding of many specialist kinds of hound, each with its own local variations – deerhounds, wolfhounds, boarhounds, foxhounds, otterhounds, bloodhounds, greyhounds and spaniels, to name but a few, although these are not necessarily the direct ancestors of the breeds that bear the same names today.

  Medieval dogs

  The mtDNA of some existing breeds shows that their identity extends back in an unbroken line at least 500 years, and possibly much longer. Some of these ancient breeds are oriental, including the shar-pei, Shiba Inu, chow-chow and Akita. Others, including the Afghan hound and saluki, have Middle Eastern origins. A third group (malamute and husky) are Arctic dogs, while an African breed, the basenji (recently confirmed from its Y-chromosome DNA as both unique and ancient), forms the fourth. Some of the North Scandinavian breeds, such as the Norwegian elkhound, have probably been derived from interbreeding wolves with dogs, several hundred and possibly as long as a few thousand years ago.

  Speciality breeds may have originally had other uses besides the standard ones, such as tracking and hunting. Several types of dog, such as the chow-chow and the fat Polynesian types, were developed specifically for food; others, such as the Manchurian long-haired types, were probably bred for their fur as well. Breeding dogs is not a particularly efficient way to obtain nourishment or something to wear, so we have to presume that there was always some social significance attached to these uses: dog meat may have been prized as a delicacy, and dog fur may have carried a higher social cachet than the hide of hunted animals such as gazelle.

  Whatever one may think of such uses for dogs, they are a testament to the dog’s extreme adaptability to the twists and turns of human civilization. Dogs have been adapted, or have adapted themselves, to all kinds of roles, in a way unmatched by any other domestic animal, and such flexibility must lie at the heart of the enduring power of the human–canine relationship. Although today most dogs are valued primarily for their companionship, at least in the West, we must also remember that historically many dogs were kept first and foremost because they were useful. Some of these functions must have come and gone in just a few centuries; just a footnote to the dog’s association with man, they are now almost forgotten (see box – ‘The Turnespete’). Others – such as hunting, shepherding and guarding – persist today.

  THE TURNESPETE

  The sole purpose of this British ‘breed’ of dog was to run in a mousewheel-like contraption, which, through a system of belts and pulleys, slowly turned a joint of meat roasting over an open fire. This device was first mentioned in the mid-sixteenth century and had disappeared by the mid-nineteenth, replaced by more efficient ways of roasting meat without burning it. Actually, purely mechanical methods of turning spits had become available in the seventeenth century – Leonardo da Vinci had sketched one – so the continued use of dogs for this purpose for a further 200 years may reflect not a strictly utilitarian consideration but a British preference for using dogs wherever they could. The dogs were certainly given names – Fuddle, one of the turnspit dogs at the Popinjay Inn in Norwich, even had a poem written in his honour. On Sundays, it was apparently the custom to take them to church where they would act as foot-warmers on cold winter days. Incidentally, there is no evidence that the Turnespete was ever a specific breed in the modern sense of a closed gene pool; short-legged and stocky, turnspit dogs were probably selected from a variety of terriers, including, according to one record, badger-hunting dogs. However, the one surviving specimen, a stuffed dog displayed at Abergavenny Museum in Wales, is more reminiscent of a dachshund.

  Modern sensibilities would be offended by such a use of dogs today. Imagine how frustrated these dogs must have felt, endlessly running nowhere while the tantalizing aroma of roasting meat was all around them. Yet their continued use even when mechanical substitutes had become available could be explained by an affectionate attitude towards these dogged little workers, rather than simply a reluctance to embrace new technology. And don’t we still give running-wheels to caged mice, hamsters and gerbils, on the grounds that they ‘need exercise’?

  European breeding restrictions were comparatively lax at first, and developed relatively late. The fact that the few genetically isolated ‘ancient’ breeds come from such far-flung locations (and none from Europe) suggests that they are relics of dogs that were carried, by human migration, out of Asia and south-eastern Europe, and were subsequently not interbred with more recent migrants, the most notable of which would have been the diverse types of dog developed in Europe in the Middle Ages and subsequently spread by colonialism. Such genetic isolation indicates a greater degree of human intervention in reproduction than for many other types of dog, although it is not possible to tell how much of this would have been achieved by selecting purebred partners for mating, and how much by culling or simple neglect of accidentally cross-bred puppies. By contrast, the DNA of modern dogs indicates that cross-breeding between different types of dog was commonplace in Europe and America. While much of this was probably accidental, historical records also indicate some deliberate breeding of unlikely combinations of types, just to see whether some useful new type might emerge.

  Aside from the few genetically isolated ‘ancient’ breeds, cross-breeding of dogs continued apace in Europe and North America until the middle of the nineteenth century. The idea that a dog should be mated only with other identical dogs is a comparatively new one, dating back only about 150 years in Europe, the same notion then spreading rapidly to other countries. Nowadays, if a dog is to be registered as a particular breed, its parents, grandparents and so on for many generations must also have been registered as the same breed – a restrict
ion known as the ‘breed barrier’. Although many mongrels and cross-bred dogs continue to be born in the West, they are much less likely than pedigree dogs are to find homes and leave offspring of their own.

  Pedigree breeding is the third phase of the transition from wolf to modern dog, each phase of which having been abetted by a different selective pressure. The first was the initial selection for tameness, from wolves that were already pre-adapted to scavenging from man. As we have seen, this process must have been essentially passive: the wolves that could tolerate interaction with man gradually isolated themselves reproductively from their wild cousins and became proto-dogs. In the second phase, deliberate selection by man for specific functions then began to become a factor, through attempted isolation of one type of dog from another. However, this was rarely, and then only locally, the factor controlling which dogs had descendants and which did not, given that deliberate selection occurred as isolated exceptions against a background of some deliberate, and much accidental, interbreeding. By contrast, the third and most recent stage of the transition from wolf to dog has seen an explosion of deliberate selection. Dogs are mated with other, virtually identical dogs, in an attempt to create ‘ideal’ breeds – most of which are cherished for their appearance, not their functionality.

  Domestication has been a long and complex process, and despite the self-evident differences between types of dog, every dog alive today is a product of this transition. What was once another one of the wild social canids – the grey wolf – has been altered radically, to the point that is has become its own unique animal. In the course of this change, the dog has shed many of its wolf-like attributes, so much so that there is no reason to presume that the characteristics that define today’s dogs are derived specifically from wolves; most of these are either products of domestication or general features of canids that predate the evolution of the grey wolf.

  Whatever the selective pressures governing them, many of the characteristics that separate domestic dogs from the wild canids can be ascribed to changes in the rates at which the body and behaviour mature. As noted earlier, dogs are in many respects similar to juvenile canids; although they grow into adults in the narrow sense that they become capable of reproducing, they also remain immature in many other respects – a sort of arrested development that neatly accounts for the way they depend on their human owners for the whole of their lives.

  Despite the differences between breeds, dogs are recognizably dogs – and not just as far as we humans are concerned. Dogs evidently recognize other dogs as such, even when the disparity in size and shape between them makes it seem implausible that they could. Dogs of all breeds, or almost all, must therefore retain some common social repertoire, enabling them both to recognize one another as dogs and to engage in some at least rudimentary communication. The question, then, is to what extent the dog’s social capabilities are a product of domestication, and what has been inherited directly from the wolf – or possibly from even further back in the canids’ evolutionary history.

  3

  Why Dogs Were – Unfortunately – Turned Back into Wolves

  Today’s dogs are clearly not wolves on the outside, but their behaviour is often interpreted as if they were still wolves on the inside. Indeed, now that we know for sure that the wolf is the dog’s only ancestor, it seems impossible to avoid such comparisons. The idea that dogs retain most of the wolf’s essential character is not only out of date, but also reflects some deep-seated misconceptions about wolf behaviour that science is only now beginning to overturn. Despite these holes in the dog-wolf theory, however, it is still widely used to inform dog training, with unfortunate consequences for dog and owner alike.

  For more than fifty years, the concept of a dog as a wolf dressed up in a cute package dominated dog training and management, with results that were – to say the least – mixed. Some parts of the advice that logically flowed from this misconception are harmless, but others, if applied rigorously, can damage the bond between dog and owner. Moreover, equating dogs with wolves allows trainers and owners to justify physical punishment of the dog, by the analogy that wolf parents achieve control of their offspring through aggression.

  The concept that dog behaviour is little changed from that of wolves does not sit well with the self-evident friendliness of the large majority of dogs. Most dogs love meeting other dogs, and most love people. This may seem a blindingly obvious statement, but, from a biologist’s perspective, it is one that demands explanation. After all, neighbouring cats often spend their whole lives avoiding one another, whereas many dogs will try to greet every dog they come across. Where does this general affability come from?

  The dog’s sociability is even more remarkable when compared to that of its ancestors. Wolves from different packs try to avoid one another; if they do meet, they almost always fight, sometimes to the death. This is not unusual; modern biologists view all co-operative behaviour as exceptional, because the default behaviour of every animal should be to defend itself and its essential resources – its food, its access to mates, its territory – against all others, and especially against members of its own species, since these must be its most direct competitors. Wolves are no exception to this rule, and any wolf that failed to compete in this way would, all other things being equal, produce fewer offspring than its neighbours. Logically, therefore, any gene that predisposes a wolf to put the interests of other wolves first should eventually disappear. Of course, kin selection means that wolf packs composed of family groups do co-operate, because this co-operation enables them to propagate their genetic material most effectively. But unrelated groups, which share far fewer genes, will either avoid one another, or fight, if they do happen to meet.

  Dogs, unlike wolves, are extraordinarily outgoing – yet even this trait has been interpreted as fitting into the idea of dogs’ underlying wolfishness. Dogs that are self-evidently unrelated – say, from different breeds – are usually perfectly happy to meet when out being exercised by their owners. Yet many old-school trainers and dog experts would argue that dogs are friendly only because they are trained to be so. Within every dog, they maintain, lurks a savage wolf that could spring at any moment for the throat of any dog it meets, unless its owner remains in vigilant control. Despite having being comprehensively discredited by biologists and veterinary behaviourists more than a quarter of a century ago, this idea continues to have a surprisingly wide currency. Many training manuals still emphasize the need for constant vigilance against the moment when young dogs begin their inexorable attempts to dominate or control all around them, dog and human alike. The only answer, they say, is to make sure from day one that dogs know that their owner is boss – a stance that humans are supposed to be able to achieve by mimicking the way that dominant wolves control their packs.

  Clearly, dogs’ easy sociability requires further examination. Inasmuch as we do not have access to the world of the ancient, tameable, wolves from which dogs are descended, perhaps it would be best to set aside the origins of the dog altogether. Put simply, the question could be: how would dogs organize their lives if they had the choice, away from mankind’s interference? Of course, this is not an easy question to answer because there are very few dogs who live free from human supervision. Dogs rarely survive for long far from human settlements, not least because domestication has all but destroyed their ability to hunt successfully. Although some elements of hunting behaviour have indeed been retained in some working breeds, few dogs, if any, possess the innate ability to put all these elements together, in order to locate, hunt, kill and consume prey on a regular basis – and certainly not when competing with other predators.

  Although it is rare to find dogs that are not controlled by humans, there are enough of them for us to begin forming a picture of what a dog-run society might look like. All over the world there are millions of dogs, generically referred to as ferals or ‘village dogs’, that are just outside man’s direct control. They live on the fringes of human society
, scavenging from rubbish-dumps and begging for handouts, but they are otherwise independent of people and certainly show no allegiance to any human owner. Such dogs are commonplace in the tropics and subtropics: they go by different names, such as the pariah or pye dogs of India, the Canaan dog from Israel, the Carolina dog from the south-eastern United States, and the basenji-like village dogs of Africa. Their DNA suggests that many are indeed native to their areas. (By contrast, that of dogs from tropical America suggests descent from escaped purebred European dogs.) There are also a number of ancient types unique to a particular location, such as the New Guinea singing dog, the kintamani from Bali, and the Australian dingo, the only completely wild dog known to have originally descended from domestic dogs.

  Urban scavengers

  Because of its uniqueness, the dingo’s story offers a tantalizing example of the social systems that dogs form when left to their own devices. Sometime between 5,000 and 3,500 years ago, a single, pregnant, domestic bitch – probably descended from the medium-sized dogs that originally evolved from wolves in Asia – arrived on the Cape York peninsula, the northernmost tip of mainland Australia, and escaped into the bush; later, her offspring were joined by a few others, transported by traders across the Torres Strait from New Guinea, as she must herself have been. On arrival in Australia, these escaped dogs found little competition from local (marsupial) carnivores, and rapidly became the dominant predator. They were thus able to, and still do, adopt several types of canid social structure; many individuals remain solitary outside of the breeding season, while others form packs of up to a dozen individuals.