The great majority of these dogs had been homed at eight weeks, which is standard because it is when the puppies can be fully weaned. However, this age is also right in the middle of the ‘sensitive period’ for socialization, so it has been suggested that a sudden change of environment at eight weeks may be particularly stressful for the puppy. There have been few other studies of this kind that could test this idea, and none in which the age at homing has been varied systematically. We therefore cannot yet be sure what is the optimum age for building up dogs’ experience of the world, although we do know that to be most effective it should start before seven weeks, and should go on for several months after that. There is also some evidence that taking puppies out of their litters before eight weeks of age predisposes them to become fearful of other dogs, so ideally the whole litter should be kept together until eight weeks, while at the same time beginning to introduce them to a wide variety of humans.
The process that leads to the bond between dog and owner is therefore set in motion when the puppy is about three weeks old. The behavioural strategies established during the sensitive period channel the puppy’s behaviour and set the ground rules for the subsequent formation of close relationships with individual human beings. Puppies that are given only very limited experience of the man-made world are likely to fail to adapt to that world when they meet it. Specifically, although they should be able to form well-balanced relationships with their owners, they may react fearfully towards other people because of their impoverished experience of the human race in all its diversity. Their default strategy for dealing with anything unknown may be to try to flee from it, rather than adopting the cautious curiosity that is the default strategy of the well-balanced dog.
All of this presupposes that a dog’s formative experience begins at three weeks of age. However, the dog has already existed for twelve weeks before this – nine as a foetus, and three as an apparently helpless puppy. When the original research into the dog’s sensitive period was conducted fifty or sixty years ago, it was thought that foetuses and helpless newborn animals were incapable of learning very much, and would grow along a predetermined path unless some catastrophe occurred. Reinforcing this idea was the anthropocentric view that newborn puppies, being blind and deaf, were incapable of learning much about their environment. Their sense of smell was largely forgotten about, although we now know that puppies can learn to distinguish between odours even before they are born as well as during their first three weeks. In short, this research largely overlooked the possibility that the twelve weeks following conception may be a time when reactions to the world are significantly influenced by outside events.
Although dogs have not been studied directly with regard to external influences during the time from conception to birth, other research indicates that this period may be particularly critical to the development of behaviour. We now know, from studies of other species, including our own, that the environment experienced by the mother can have profound effects on the character of her offspring. Research on rats, mice, monkeys and humans has shown that the development of the foetal brain can be powerfully influenced by the mother’s experiences during her pregnancy. There is no reason to suppose that the same does not apply to dogs. Most of the research on brain development in the foetus has focused on severe stress experienced by the mother. In our own species, it is now well established that maternal stress can be linked to a whole range of mental disorders in children, including chronic anxiety, attention deficit hyperactivity disorder (ADHD) and inadequate social behaviour. Longer-term problems can result, including poor intellectual and language skills, lack of emotional control, and even schizophrenia. Studies of rats have shown that these problems almost certainly stem from the effects of stress hormones produced by the mother that cross over the placenta into the foetus itself. There, they change the way that the brain is developing, resulting later, after birth, in reduced activity of some neurohormones (such as dopamine and serotonin) and a hyperactive stress-response system. Although the details vary slightly depending upon the stage of pregnancy during which the stress occurs, the young animal can then exhibit impaired learning, poor play skills, and a weakened ability to cope with challenges.
Luckily, these deficits appear to be reversible if the infant animal receives additional maternal care after birth,11 but they can also be made worse if the young animal is taken away from its mother prematurely, or if her maternal skills are deficient. Surprisingly little attention has been paid to this phenomenon in domesticated animals, despite the potential lessons to be learned about how best to look after them. One study has shown, however, that domestic sows kept in unstable social groups not only become stressed, but their female offspring are, in turn, more than usually aggressive to their own piglets, implying a profound and long-lasting effect on the development of their brains.12
Are these effects all pathologies, or do some of them actually prepare the young animal to cope with a changing world? Because much of this research has been done in an effort to substantiate factors affecting mental illness in humans, less thought has gone into trying to understand why evolution has allowed stress experienced by the mother to affect her offspring so profoundly. Indeed, the general assumption has been that these are indeed pathologies, beyond the reach of natural selection. Nevertheless, research on guinea pigs (and birds) suggests that it may sometimes be helpful for offspring to be pre-programmed in this way. Female guinea pigs that give birth in overcrowded social groups tend to produce female pups that behave more aggressively than normal. Their male offspring, on the other hand, become ‘infantilized’, for example continuing to play-fight at an age when their normal counterparts are competing with other males for real. These changes may actually prepare the young guinea pigs for the environment in which they will live; in order to find food and space to breed in a crowd, females need to be pushy, whereas young males need to keep their natural competitive nature in check until they are big and strong enough to win against the most experienced males.
It is quite possible that some of the changes in the brain wrought by maternal stress are, at least to some extent, adaptive, in the sense that they prepare the offspring for an uncertain world. However, this is more likely to be the case with wild animals than with domestic animals such as dogs; any response to stress that evolved in the wild ancestor, the wolf, is unlikely to still be adaptive in the man-made environment of today.
All these findings strongly suggest that dog breeders should place great emphasis on the psychological well-being of their breeding bitches. They should neither induce the stress of separation by isolating them for long periods, nor allow the bitches to be intimidated by other dogs. Some of the deficits that we found in dogs born in non-domestic environments were quite possibly due as much to stress induced in their mother as to impoverished experience during the first eight weeks of their lives. (The owners who brought their dogs in for treatment were simply unable to give us this much detail about the environment where their dogs had been born.) Purchasers of puppies would therefore do well to examine the conditions under which breeders keep their bitches, as well as the environment experienced by the puppies themselves. Of course, owners also have a duty to ensure that their new puppy gets the right kind of experiences during its first few months, but however good these are, there is a risk that they may not be sufficient to completely reverse the consequences of having had a chronically stressed mother, or an impoverished environment during its first eight weeks.
Overall, it is clear that a puppy’s experiences, from soon after conception to when it is roughly four months old, play a crucial role in affecting its character. A dog that gets the wrong start in life can grow up to be overly fearful or anxious. This is not absolutely inevitable, as nature has built in a capacity, up to a point, to compensate for setbacks early on and return development to a balanced trajectory. Nevertheless, there is much we still do not know about why some dogs develop behavioural problems and others do not.
Most dogs seem to become more distressed when they are separated from their owners than when they are separated from other dogs. So the question arises, do dogs love people more than they love other dogs? This does not sound like a particularly scientific sort of question, but it could be a test of just how domesticated dogs have become. Few scientists have ever considered this a question worthy of an answer, but there is one study that conclusively shows that dogs are indeed prone to bonding more strongly with people than with other dogs.13 The subjects of the study were eight seven-to-nine-year-old mongrels that had been living as littermate pairs in kennels since they were eight weeks old; all had been fully socialized to people, and they were being looked after by one carer who was, as far as they were concerned, equivalent to their ‘owner’. When the experiment began, the kennelmates had not been apart even for a minute during the previous two years, and hardly ever during their whole lifetimes. However, when one of each pair was taken out of earshot for four hours, the remaining dog’s behaviour did not alter appreciably. Puppies separated from their littermates will usually yelp until they are reunited, but these adult dogs barely even barked. Moreover, the level of the stress hormone cortisol in their blood did not change as a result of the separation, provided the dogs had been left in their familiar pen. Overall, therefore, there was no indication that any of these dogs was upset, this despite the fact that, since they had virtually no history of being left alone, they would not have been sure that they would be back with their pen-mate in a few hours’ time.
In contrast, when the dogs were taken to an unfamiliar kennel, they did become upset. They were visibly agitated, and their levels of stress hormone went up by over 50 per cent. Remarkably, this proved true whether they were on their own or with their kennelmate. When the two were together, they did not interact with one another any more frequently than usual; whatever the bond between them, it was not sufficiently comforting or confidence-building to help them cope with being somewhere new, outside their familiar territory. However, if their carer sat quietly with each dog in the novel kennel, it would stay near him and pester him for contact (which he responded to by brief episodes of stroking). This was apparently enough to alleviate the dogs’ stress completely, because if the carer was there, their cortisol levels stayed close to normal.14
These dogs, although they had kept the company of another dog for their whole lives, behaved as if they were much more attached to their carer than to their brother or sister. While they had not led quite the same kind of life as a pet does, everyday experience suggests that the same is probably true of pet dogs. Dogs do have territories, in the sense that they feel most calm when they are in familiar places, but, like the wolf, they can comfortably go to new places if they are with their ‘pack’, the difference being that in this case the key ‘pack’ member is almost always a human (namely, the owner), and not a member of their own species. For many dogs, the owner will be a constant feature of their lives from the middle of the socialization period onwards. However, others will be forced, through changes in circumstances, to alter their primary attachments on several occasions during their lifetime. So in addition to the capacity to accept both humans as well as dogs as social partners, domestication has given dogs sufficient social flexibility to form new ‘familial’ ties at almost any time in their lives.
Since the need for a human attachment figure seems to be unusually powerful in the domestic dog, dogs that are abandoned by their owners and end up in rehoming centres must feel this very acutely. Research has shown that just a few minutes of friendly attention from one person on two consecutive days is enough to make some of these unowned dogs desperate to stay with that person; when left on their own, these dogs will howl, scratch at the door that the person has left through, or jump up at the window to try to see where she has gone. For many dogs, this perception of humans as potential attachment figures may last their whole lives; luckily for many of them, one individual or one family will satisfy this need from eight weeks of age for the rest of their lives. This craving certainly explains why so many dogs develop separation disorders at some point in their lives.
Even though most of the evidence for these strong and rapidly forming attachments comes from the behaviour of dogs that are distressed by separation, the strength of such attachments suggests that they should also be expressed in pet dogs’ normal behaviour. Unfortunately, however, very few biologists have studied the everyday interactions between pet dogs and the families they live in. There are probably a variety of reasons for this: such studies are time-consuming; they use techniques more commonly employed by anthropologists, who are rarely interested in animals; the data they generate are complex and not straightforward to analyse; and there is the real risk that the mere presence of an observer would change the way that family members behave towards their pet. For example, some people may feel inhibited, while others might use the opportunity to ‘show off’. Nevertheless, such studies are a very useful counterpart to the much more structured investigations of, for example, dogs’ cognitive abilities.
One of the earliest, and still one of the best, of these ethnographic studies shows just how people-focused most pet dogs are. Ten middle-class dog-owning families living in the suburbs of Philadelphia were observed for a total of twenty to thirty hours, usually in the late afternoon and early evening when the children were at home. The researcher noted that the dogs paid much more attention to the human members of the household than vice versa. They watched, approached or followed one or more household members. When they rested, they often faced people in the same or the next room. When they happened to be looking elsewhere, such as out of a window, they were evidently still aware of where the people were, often turning towards them and approaching. Conversely, however, the family members rarely interrupted what they were doing to seek out the dog when the dog was in another room.15
The apparently single-minded vigilance on the part of the dog was not uniformly directed, however. Dogs are very good at sensing who in the family likes them best. In the three families where the husband was not attached to or interested in the dog, the dog seldom watched or followed him. In this way, the dogs showed that they had learned who had been most responsive to them in the past. The implication is that dogs, once they have an attachment figure, are not indiscriminate in terms of who else they become attached to, presumably relying on their experiences of people to guide them in how they should best react.
The ‘lupomorph’ or ‘pack’ model, while flawed in many respects, is therefore correct in one: dogs do behave towards us using a set of rules and behaviour patterns that are ultimately derived from those of the wolf and more distant canid ancestors. However, these are not the rules of ‘Dominate or be dominated, crush or be crushed’. They are the rules of the family, the rules of ‘Those who raise you are those who are most likely to continue to cherish you throughout your life’. Our pet dogs’ behaviour clearly shows us that they see us as attachment figures, based on a parent–offspring framework. Indeed, the dog owner who tells her friends ‘I’m Fido’s mum’ is really not far wrong.
Dogs are fundamentally different from all other animals in this respect. We take it for granted that we can exercise them off-lead, and that, once trained, they will return to us for no more immediate reward than being reunited with us. The mechanisms involved are essentially developmental; domestication has imbued the dog with the capacity to achieve this unique social behaviour, but it is only through the learning environment we give them that dogs come to un
derstand how to behave towards people.
6
Does Your Dog Love You?
Dogs are obviously attached to their owners – in the sense of their behaviour, in the sense that they follow them around. But does your dog actually love you? Of course it does! It tells you, every time you come home, by the way it greets you. Your dog may be ‘just’ a household pet, but I would be very surprised if most owners could not bring themselves to say that they loved their dog, and that their dog loved them in return. Anything less, and the relationship is probably in trouble.
Emotions are not easy to pin down, scientifically speaking. As a scientist, I can investigate how much you love your dog, and, as a human, I can be reasonably sure that what you describe to me as ‘love’ is much the same emotion that I have felt for my own dogs. We can both articulate this, first because we are members of the same species and therefore are likely to have similar emotional repertoires and, second, because we can communicate our feelings to each other through language.
However, the love that flows in the other direction, dog to owner, is much harder to pin down. First of all, dogs cannot tell us how they feel, so we have to deduce it from their behaviour. Can we be sure that we always get this right? Second, because we belong to different species, we cannot simply assume that dogs experience the same array of emotions that we do. In fact, I would go so far as to say that it is unethical to make that assumption. Scientists have a responsibility to convey as much as they know about the reality of canine emotions, guiding owners to a proper perception of what their dogs can and cannot feel.
In Defence of Dogs Page 18