In Defence of Dogs

by John Bradshaw

  Probably the earliest well-established archaeological example of a dog that is both biologically distinct from wolves and specially connected to humans is the burial, about 12,000 years ago, in what is now northern Israel, of a human with one hand resting on the body of a puppy. Not only does the position of the puppy show that it had a close relationship with that person, but its teeth are also significantly smaller than those of any wolf that lived nearby at that time, indicating that it must have come from domestic stock.

  Neither the physical signs of domestication in this puppy, so distinct from its wild counterpart, nor the evident bond between the animal and its owner, can have arisen overnight. Rather, the puppy must have been preceded by many generations of dogs that made up the transition from wild wolf to domesticated pet. Such transitions may be virtually invisible to archaeology, but the subsequent rapid emergence of dogs all over the Old World is compatible with the idea that there was not one domestication, but several. In the next two millennia after this 12,000-year-old burial, other similar burials – either of humans and dogs together, or of dogs on their own – occurred in various parts of Europe. These sites have been found in, among other places, the United Kingdom, suggesting that dogs had also spread quickly from their points of origin, which are thought to be far to the east. It is thought that, at roughly the same time, humans were taking other domestic dogs, probably from another focus of domestication in East Asia, out of Siberia and into what is now Alaska. (At the time, both were part of a single land mass, known as Beringia, which, depending on the period, stretched as far as 600 miles from north to south.) These dogs moved with one of the early waves of colonists down the west coast of North America, and then into the interior: the earliest confirmed dog remains in the Americas, in Danger Cave, Utah, are perhaps 10,000 years old. Meanwhile halfway across the world, humans took dogs with them as they moved into the furthest reaches of south-east Asia; the DNA of the 800,000 street dogs found in Bali today shows that they are the descendants of dogs that arrived there overland, before Bali became an island 12,000 years ago.

  This rather rapid appearance, in the archaeological record, of dogs all over the globe can, at a stretch, be explained by many independent domestications taking place almost simultaneously – but it is also plausible that domestication of the dog did actually start much earlier than the archaeology indicates. The point at which archaeologists can be sure that the dog had already become a domesticated animal may actually reflect not the beginning of the transition from the wolf, but rather the culmination of a fundamental change in the relationship between man and dog, one that had already taken thousands of years to develop. This process could not be complete until the dog had become an integral part of human culture, and also until it no longer needed to maintain the physiognomy of the wolf, because many of its essential needs were being taken care of by its owner. Thus the 5,000-year discrepancy between the date of domestication as shown by the archaeological record and that indicated by the dog’s DNA can be accounted for by positing an extended period over which domestication took place gradually. These earliest dogs, or proto-dogs as they are sometimes called, would have been indistinguishable from wolves in the way they looked, and were probably treated in a strictly utilitarian way. For example, they may have been communal ‘property’, as today’s village dogs are, rather than having a single ‘owner’.

  To be sure, any pre-domestication theory that suggests several thousand years of coexistence between wolves and people before the transformation to domestic dogs must account for the lack of archaeological evidence over this period, say from 20,000 to 15,000 years ago. If dogs existed during this period, possibly even earlier, why are they absent from human burials for the whole of this period, but then suddenly start appearing in burials, all over the globe, over the course of ‘only’ a couple of millennia? The archaeological record itself may hold the answer.

  The earliest known dog burial is more than 14,000 years old. Located in Bonn-Oberkassel in Germany, it was discovered in a quarry in 1914, and seems to have consisted of a partial skeleton of a dog, buried alongside two humans. Unfortunately the outbreak of the First World War led to the loss of much of this material; yet a single piece of the dog’s jaw remains, the arrangement of its teeth showing that it was clearly not from a wolf. Archaeological evidence suggests that from then on, dog burials become almost commonplace (other kinds of animals were occasionally treated in this way, but nowhere near as frequently as dogs2). Some dogs were buried alongside people, others had their own dedicated graveyards. In what is now the south-eastern United States, dog burials were so common during the period between 9,000 and 3,000 years ago, that it is their relative infrequency from later burial grounds that archaeologists feel they need to explain, rather than the other way around.

  Mankind had been burying its dead for tens of thousands of years before dog burials began. Many ancient human graves contain animal remains; some may have come to be there accidentally, but many were obviously included deliberately, indicating a powerful emotional link between early humans and the animals they found around them. Consider this description of the contents of a grave, dug 28,000 years ago in Russia, that contained the remains of a boy, a girl and a sixty-year-old man. Buried with them were thousands of pieces of deer’s antlers, polar foxes’ teeth and mammoth ivory, which had probably been incorporated into necklaces or as decoration on their long-disintegrated clothes. Alongside the boy was a sculpture of a mammoth, itself carved from mammoth ivory. In another grave nearby there was a small ivory sculpture of a horse (a hunted animal at this point, not domesticated). These people had clearly had an important relationship with their local animals, one that included representing them in their art, and possibly featuring them in their religion. This relationship, however, seems to have been exclusively that of hunter and quarry.

  The absence of dogs in known burial sites older than 14,000 years almost certainly means that dogs were, before then, rather rare. If the culture represented in this group grave in Russia had used dogs for hunting, it seems likely that there would also be evidence of dogs in this or similar graves – either bones, or, as with the horse, some kind of representation. That there are no such traces indicates that the society from which these people came did not have domestic dogs. If they had, these dogs’ remains would have probably been indistinguishable from those of wolves, but in fact there is no trace of any wolf-like animal, domestic or wild, in the Russian group grave, even though wild wolves would have almost certainly have been in the vicinity. Indeed, very few graves hold traces of ancient wolves.

  Unlike dog burials, which as noted were quite common after they first appeared in the archaeological record, wolf burials – whether alone or accompanying human burials – seem to have been extremely rare throughout ancient human history. (If common, they might have provided evidence for the early stages of domestication, when the bones of proto-dogs would have been indistinguishable from those of wolves.) Wolf teeth do feature, alongside those of other predators, in many burials of humans, but their significance is usually unclear, and in any case many probably come from animals that were killed for their pelts. The close emotional relationship that hunter-gatherers evidently had with the animals they hunted seems not to have been extended to their competitors in the hunt, including wolves. Thus there is very little archaeological evidence indicating any kind of relationship between hunter-gatherer man and wolves, either wild animals or those already on the path to domestication, until dogs suddenly appear in burials, some 14,000 years ago.

  A wolf burial from near Lake Baikal in Siberia: its limbs enfold a human skull

  Among the very few wolf burials that have been discovered, one is particularly odd, and may provide evidence for the transition from wolf to dog. Russian archaeologists recently discovered, in a cemetery near Lake Baikal, what they have identified as a wolf, buried with a human skull between its paws. The burial probably dates from only about 7,500 years ago, by which time there may well
have been dogs in this area. What is remarkable about this wolf is that it was not local; it appears to be a tundra wolf, and if so must have travelled several thousand miles before ending its days in this grave. But what if the animal is not a wolf at all? Rather than a far-flung tundra wolf, the wolf buried near Lake Baikal is, I believe, more plausibly the descendant of a ‘socializable’ tundra wolf that had been adopted, many generations before, as a pet. Under this interpretation, such a burial gives us a tantalizing glimpse of the process of domestication in action. Domestication is very unlikely to have involved a steady transition from one group of wolves to today’s dogs; on the contrary, it was more likely to have been a more haphazard process in which several domestications occurred, in different places and at different times. The ‘wolf’ in the grave may in fact be a proto-dog, the product of a late attempt at domestication in the frozen north, brought south, where it lived and died alongside its more ‘domesticated’ cousins – the progeny of earlier domestications – that were, by then, recognizable as dogs.

  Archaeologists have found a few other burials of ‘wolves’ that may in fact be proto-dogs. For example, 8,500 years ago in what is now Serbia, a small type of domestic dog was used for food, as attested by the many broken leg-bones and skulls found in rubbish pits there. Another (larger) type of dog from the same area and at about the same time was buried unharmed in proper graves, implying a role that included companionship. Even more pertinent, however, is evidence – from the same location and period – of the remains of what appear to be wolves. These may have been wild, but it is also possible that they were a third type of dog, which, unlike the other two, had not changed much in appearance from its wild ancestor.

  Few traces of proto-dogs have been found in human burials, but is there anything in the fossil record to support the idea of gradual and haphazard domestication? Until recently, archaeologists were reluctant to identify wolf skulls more than 14,000 years old as belonging to anything other than a wolf, so any possible proto-dogs that were found were not labelled as dogs. The earliest distinct dog skull, from Eliseevich, in the Russian plain, was excavated from the edge of a pile of mammoth skulls, and it, too, has been dated to about 14,000 years ago; roughly the size of a husky’s skull, its burial seems to have been accidental rather than deliberate. However, three new skulls have recently been found that are intermediates between those of wolves and early dogs, such as the Eliseevich dog. All three are very similar to today’s Central Asian shepherd dog. The oldest of these skulls, from Goyet in Belgium, is a staggering 31,000 years old, more than twice the age of the oldest dog burial. The other two, from Ukraine, are probably only about 13,000 years old, roughly contemporary with the first dog burials. The Goyet specimen is therefore something of an anomaly. Could it have been a direct ancestor of today’s dogs? Or is it our only record of a very early domestication of the wolf – one that failed, hence the absence of any trace of dogs for the next 17,000 years?

  There is one more small piece of evidence that suggests a relationship between man and dog going back more than 20,000 years. Deep in the Chauvet cave in the Ardèche region of France, which is famous for its prehistoric art, a 50-yard trail of footprints made by an eight-to-ten-year-old boy, alongside those of a large canid, hints at a close relationship between the two. The footprints of the canid are intermediate between those of a dog and a wolf. Soot from the torch that the child was carrying dates the event at 26,000 years ago, making these probably the oldest human footprints in Europe. With a little imagination, one can picture a boy and his faithful (proto-)hound, venturing into the cave to view the spectacular representations of wild animals, painted on its walls 5,000 years previously.

  Single footprints of a child and a canid in the Chauvet cave, in the Ardèche region of France

  Evidence such as the foregoing is, ultimately, too flimsy to give us any firm idea of when or where domestication of the wolf began. Nevertheless, this process does appear to have been repeated several times, in several locations in Europe and Asia, over a period of many thousands of years, to the point where domestic dogs may already have been established in some parts of the world at the same time that wolves were being taken from the wild in others. Some of these attempts must have succeeded; others almost certainly failed, leaving no trace in today’s dogs. The habit of burying dogs with humans seems, for some reason that is still unclear, to have been adopted only after domestication was well advanced, otherwise there would be human graves from somewhere between 25,000 and 15,000 years ago that contained the bones of proto-dogs, indistinguishable from those of wolves.

  We can say for certain, however, that the earliest confirmed dogs, at 14,000 years ago, logically represent not the beginning of domestication, but rather the end of its first phase, marking the point at which dogs became physically distinct from wolves. Before that, there must have been changes in the brains of wolves that made them suited to living with people but which left little or no trace in their skulls for archaeologists to find today. The question still remains as to how long those changes took, how many of those first alliances between dogs and wolves failed, and how many leave their traces in the dogs of today.

  Since we can account for the diversity of the dog’s DNA only by hypothesizing that individual domestications of wolves occurred in different parts of the world, these different ‘proto-dog’ populations must initially have been isolated from one another, and probably remained so for thousands of years. As domestication progressed, however, these early dogs would have eventually become manageable enough to travel with people on large-scale migrations, thereby allowing individuals from one proto-dog population to meet up with, and begin interbreeding with, individuals from another. The resultant churning of the dog’s gene pool probably started more than 10,000 years ago, such that even dog remains that are old enough to be fossilized may have originated in wolf populations many hundreds or even thousands of miles away.

  Thanks to this complicated timeline of domestication, the location of the original domestication events has proved impossible to pin down. The wolf itself is a highly mobile animal, even though it has not benefited from being transported around by man. Migrations of wolves, even since the dog was domesticated, have resulted in the incidence of almost identical DNA among individuals in regions as far apart as China and Saudi Arabia. Thus the DNA of modern wolves gives only very weak clues as to where domestication might have taken place.

  Leaving the wolf to one side and approaching the locational problem from the opposite end, other biologists have recently analysed the DNA of local ‘village dogs’. The scientists’ hope was that these dogs would be the direct descendants of the first wolves to be domesticated in the area, and the assumption was that, as dogs dependent on humans, they were much less likely than wolves to have travelled long distances since then.3 One recent study has suggested that, because the village dogs of southern China have the most varied DNA found so far, that was where domestication must have occurred – but since then it has been discovered that village dogs are almost as diverse in Namibia, where the nearest wild wolf is 3,000 miles away.4 To have got that far, and to have become so widespread (there is little difference between the DNA of village dogs in Namibia and Uganda), these village dogs must have had considerable help from mankind; perhaps they accompanied humans on their various migrations around Africa. There also appears to have been a substantial amount of interbreeding between apparently localized village dog populations, resulting in a gradual trickling of greater diversity into their DNA – even in places as isolated as Namibia.

  Village dogs

  Despite all this considerable – and continuing – research, there are still no firm answers to the question of where the dog was domesticated. It must have happened in areas where wolves occurred naturally, yet North America has been ruled out, since the DNA of North American wolves is quite different from that of domestic dogs. This leaves most of Europe and Asia as possible locations. Beyond this consensus, there is a great de
al of conjecture, even disagreement, among the various researchers pursuing a definitive answer.5

  The most likely scenario is that wolves were domesticated at several different locations, possibly across Asia, including the Middle East, although one or more additional European origins also seem plausible. Taken at face value, the archaeological evidence points to at least one early origin in the Fertile Crescent, and this is the preferred scenario of some of the DNA experts. However, one interpretation of the DNA points to the earliest domestication as occurring in South China, where far less archaeological investigation has taken place. Each of the teams of DNA experts have their own samples of dogs and wolves, and so far these have led them to different conclusions. Their accounts are not easy to reconcile at present, but the most likely conclusion is that there was no single point of origin, but rather that wolves entered human society at several far-flung locations across Asia and Europe. Some left few or no descendants; others prospered, and eventually interbred as humans began to take dogs with them as they travelled.

  Although we still do not know where, exactly, the dog originated, it is clear that our modern dogs do not trace their ancestry back to any one particular kind of wolf. Dogs are the result of a mixing of many different kinds of wolves from across Asia and Europe; the only wolf that is definitely missing from the recipe is the American timber wolf. Thus there is no wolf alive today that can act as the perfect model for understanding dogs and the way they behave. Moreover, the long period over which domestication occurred means that dogs have had the opportunity to change radically since they became separate from wolves, 10,000 or more generations ago. During the same period their environment, too, has undergone a considerable transformation.