Q&A with Dog Sense Author John Bradshaw

What is an anthrozoologist, anyway? Turns out it’s someone who studies human-animal interactions, and John Bradshaw, who directs the world-renowned Anthrozoology Institute based at the UK’s University of Bristol (and founded it at the University of Southampton), is pre-eminent among them. For more than a quarter of a century, he’s investigated the behavior of dogs and their people, and his findings have been widely published. In Dog Sense — his best-selling, recently released book — he expands upon his belief that “the future of the dog does not lie simply with the blunt instruments of legislation and regulation, but with better public understanding of what dogs actually are, their needs and wants.” Recently, Bradshaw shared his thoughts on evolution, training (debunking the myth behind the “dog as wolf” model), changes in breeding practices in the UK and what lies behind dogs’ attraction and attachment to us, among other intriguing ideas.

Bark: Why do you think that a proto-dog — a transition from wolf to dog — evolved?

John Bradshaw: My theory — and I have nothing to back it up — is that something happened in the brains of certain wolves that made dual socialization possible. Humans developed a propensity to take in pets, and then these particular wolves came along — these would be the protodogs. They would have looked exactly like wolves. This was not an intervention on our part, but rather, a very different cultural environment.
A key difference between dogs and wolves is not their appearance but rather, how they behave. Dogs have the capacity to socialize to both species, ours and their own, and the unique ability to continue functioning as members of their own species while simultaneously establishing and maintaining relationships with ours.

B: Most researchers refer to domestication as a one-way street. Didn’t other species, including the wolf and proto-dog, also have an effect on our own evolution?

JB: Domestication was a long and complex process; speculatively, I would [say] that there were several failed attempts. Researchers who are studying human evolution and the human brain pretty much say that our own evolution — at the genetic level — wasn’t influenced by dogs. But, of course, our culture has been profoundly influenced by them.
Dogs were, for a long time, a crucial part of our technology and their domestication marked a technological innovation that also provided the blueprint for the domestication of other animals; if we were able to domesticate dogs, why not pigs, sheep, cattle, goats? So if you are talking about evolution in the general sense of where humans are today, what we think about and how we see the world, then, yes, dogs dramatically affected that evolution. If you are talking about dogs affecting genetic evolution, we haven’t discovered that yet. I’m not saying we won’t, but we aren’t there yet.

B: Do you think it’s possible that we hunted together, or perhaps learned or honed our own skills by watching wolves hunt?

JB: I don’t think we were hunting partners, to begin with, but one of the versions of human evolution that I strongly subscribe to comes from Steven Mithen, a cognitive archaeologist and professor of early prehistory, who studies the evolution of the human mind and why we are different from the Neanderthal — why they died out and we didn’t. One of the key [dissimilarities] he points to is our ancestors’ ability to think like animals. They could put themselves in the place of an animal — that they, in fact, had a connection to the animals. So we would be able to think, “If I were a wolf, what would I be doing?” or, “If I were a deer, what would I do now?”

B: If scientists have concluded that wolf behavior is different from that of dogs, why do people still consider the lupomorph (wolf pack) model as a determinant of canine behavior?

JB: They have a good excuse, which is that in terms of their DNA, dogs and wolves are so similar. However, that doesn’t mean there is similarity in their behaviors.
Confusion about how wolves actually behave comes from observations of wolves artificially grouped in zoos. A natural pack is based on a family, but those confined in zoos and so forth are not family units. So in a zoo their behavior looks like it is one of dominance hierarchy based on aggression. The whole basis of wolf behavior [in that context] is not natural. It’s like comparing all human behavior to the behavior of humans in refugee camps. In that kind of group, behavior is distorted.
The second reason is that proto-dogs, the wolves who became domesticated, were different than other wolves. The animal who was the common ancestor of wolves and protodogs has been extinct for at least 15,000 years. Wolves in the wild are getting wilder and wilder for at least 15,000 years, probably longer.
Recent interpretations of wolf behavior have emphasized cohesive, rather than aggressive, behavior as being essential to the stability of a pack. Wolves in different packs try to avoid one another, but dogs are extraordinarily outgoing. Dogs’ sociability is even more remarkable when compared to that of their ancestors.

B: If the wolf model isn’t appropriate, what is?

JB: The behavior of feral, or village, dogs in Italy, Russia and India has been studied recently, and results show that those dogs are much closer to the ancestors of pet dogs than wolves are. These are urban feral dogs, high-density dogs, dogs in large groups. Earlier studies [of feral dogs] were conducted in environments in which the dogs were being persecuted and are like the early captive-wolf studies: not reliable.
Research recently conducted in West Bengal (where feral dogs are more tolerated by the people) has found that feral dogs are a lot more tolerant of one another than wolves are. Family bonds form, but with less correlation. They do not hunt together, but rather, forage singly, and, unlike in a wolf pack, more than one female in a social group will breed at the same time. They aren’t a pack in the wolf sense; their “pack” structure is very loose and rarely involves cooperative behavior, either in raising young or obtaining food.
The studies of West Bengal feral dogs don’t offer the slightest shred of evidence that they are constantly motivated to assume leadership of the pack within which they live, as the old-fashioned wolf-pack theory would have it.

B: You write that there is little evidence that hierarchy is a particular fixation of dogs — that dogs do not want to dominate us — but so many trainers (including Cesar Millan, as you note in the book) and others use this construct to explain dog behavior. Why is this wrong and what are its implications?

JB: Part of the problem is that confrontation makes good television, and attracts programmers, but having a confrontation in your living room with your own dog isn’t the best way to train a dog. The more effective way is to use reward-based training, which can be (by television standards) incredibly dull, since it may take hours or sometimes weeks. My colleagues and I are appalled by the popularity of this style of confrontational dog training. I don’t know what the situation is in your country, but in the UK, we have a new Animal Welfare Act, and that kind of training goes against its recommendations. The law reads, “All dogs should be trained to behave well, ideally from a very young age. Only use positive reward-based training. Avoid harsh, potentially painful or frightening training methods.”
There is little evidence that hierarchy is a particular fixation of dogs, either in their relationship with other dogs or in those with their owners.
And if some trainers believe that dogs only perceive us as if we were other dogs (or wolves), there is no logical basis for assuming that dogs [instinctively] want to control us. Domestication should have favored exactly the opposite: dogs who passionately want us to control them.

B: Have you seen any changes in breeding practices in the UK as a result of the BBC’s “Pedigree Dogs Exposed” documentary?

JB: The genetic isolation of breeds has brought about a dramatic change in the canine gene pool. Three inquiries have been commissioned: one by the Royal Society for the Prevention of Cruelty to Animals, another by the government and a third by the Kennel Club itself, but there is still a great deal to be done. There are problems implementing the studies’ conclusions because the KC, like the AKC, is a federal structure made up of individual breed clubs. The federation has no power to tell the member breed clubs what to do.
There is also an unfortunate loophole in the UK legislation, in a macabre sort of way: the law doesn’t apply to fetuses so if there is a hereditary defect, it can be legal!
Top breeders, those who show their dogs, practice selective breeding to meet the latest interpretation of the breed standard, which is based on the appearance of the dog. The whole basis of judging rests on how a dog looks and behaves in the show ring.
Some of breeds’ gene pools are too small, and the answer has to be to amalgamate breeds to increase genetic variation. A group of people in Australia are taking on the breeding of pet-quality dogs, [selecting for] calm personality, trainability, freedom from inheritable disease and discomfort, people-focused and so forth. Dr. Paul McGreevy and Pauline Bennett are part of this group. Genetics can only go so far, though. You have to mold a dog’s personality — it can’t be done through genetics alone.

B: Many people use puppy testing to predict a dog’s adult character. Do you feel this is valid?

JB: Dogs are born to become friendly toward people, a process that starts in about the third week of their life and goes on for several months. This process of socialization is well charted. At 16 weeks, the window of socialization to people begins to close, though it stays open a bit longer for socialization to other dogs.
Young puppies try out different behavioral approaches; they change from one day to the next. It is more important to look at the litter’s environment — how is the female kept, for example? Puppy tests carried out at seven or eight weeks of age are being conducted when a puppy’s behavior is actually most malleable. Numerous scientific studies have failed to find any validity in puppy testing as a predicator of future character. The only personality trait that seems to be resistant to change after seven weeks is extreme fearfulness.

B: You write that dogs have been so heavily selected to form strong attachments to humans that many suffer from separation anxiety — up to 50 percent of Labs bred in the UK, for instance. On what is this finding based?

JB: It comes from my own research and that of others. We concluded that many dogs experience this anxiety at some time in their lifetime. In one longitudinal study, we followed puppies, 40 in all, litters of Labradors and Border Collies, from eight weeks to 18 months old. Over 50 percent of the Labs and almost half of the Collies showed some kind of separation distress. Subsequent studies, during which we filmed dogs left alone, showed that self-reporting by owners underestimates the scope of the problem.
We work closely with rehoming charities, instructing them on prevention and ways to train dogs so they won’t suffer when left alone. The key thing is to get new owners to train the dog to understand that they are coming back.
This is not a disorder at all, but rather, a perfectly natural behavior. We have selected dogs to be highly dependent on us. Research has shown that just a few minutes of friendly attention from one person on two consecutive days is enough to make some dogs in shelters desperate to stay with that person. Their attachment to humans is that strong.

B: One of the most controversial positions you take is that being in a shelter may damage a dog. Was consideration given to contributing factors such as the length of time spent in a shelter, the condition of the facility, the interactions a dog has with other dogs and humans there, and the dog’s personality and history?

JB: We want to understand what is going on inside these dogs, and I am not in any way blaming rescuers or shelters. Dogs who have been attached to a family may suddenly wind up in a shelter for a variety of reasons: family breakup, job loss or the dog’s behavioral problems. Dogs will be very upset by this and when they arrive in a shelter, their cortisol level [a stress-related hormone] goes sky high. We know this because when we’ve taken urine samples, we’ve had to dilute the urine to even get a measurement — it was that high. They don’t have the resources to cope and go into hyperdrive, desperate to please people. As a result, in a shelter setting, dogs actually can be easily trained.
As I mentioned, attachment can happen quickly in shelters. Of course, when dogs are unhappy, they need to be appropriately cared for, but we find that it’s important to rotate their caregivers so they don’t form an attachment to any one person.
It is also important to assess dogs for separation anxiety, predict the behavior, and advise [shelter staff and prospective adopters] on how to train them to be left alone. That is one of the most important things you can do to ensure the welfare of the dog [in terms of his or her eventual placement] in a new home.

B: Dogs clearly love us, and demonstrate that in many ways, but is this what motivates them to obey us and follow our lead?

JB: Human contact has a high-level reward value for dogs; simple attention from us is rewarding. And if that attention comes while playing with them, it can be a double reward. You can train a dog with a tennis ball, but while the game is important, it is not the only thing. The real treat is the interaction. Withdraw your attention, ignore the dog, and the dog will find this withdrawal of attention aversive.