Dogs and their people

I always watch the news after dinner, but once we’d had Ghost for a while, there was another reason to appreciate this habit. He took to lying on the other side of the room where he’d look at me. Before when I’d tried to make eye contact, his eyes would swivel away. But that changed. He’d watch me watching TV, and if I looked in his ice blue eyes he just looked right back. And his mouth was open in a lovely relaxed smile. I felt like he loved me. I’d say, “I love you!” and he’d give a little “woo!” back. 

“I love you!” 

“Woo!” 

Surely he loved me. 

Bodger alternates between nudging me or my husband with his head to demand to be petted. In his own insecure way, he loves us too. 

Of course, Ghost was responding to the tone of voice rather than the meaning. When I said something he definitely understood—“Would you like to go for a walk?”—the response was far more excited. “Woo! WOO WOO WOOOOOO!!!” He sounded like Chewbacca. 

But do dogs really love us? What do we humans mean to them? There are several strands of research that look at this question, although scientifically it’s not phrased in terms of love, which is a subjective experience. Instead, scientists have borrowed techniques from psychological studies with children and from neuroscience to investigate whether dogs have attachments to their humans. 

Dogs and Attachment to Humans 

Although all dog owners know subjectively of their importance to their pet, there’s a scientific perspective on this too. In psychology, attachment refers to a child’s close emotional bond with their caregiver and was first described by John Bowlby. Attachment is not just friendly behavior, because it includes wanting to be close to the caregiver in response to distressing events. In human children, there are four components to attachment: (1) distress when separated from the caregiver emerges at 7 to 9 months, (2) seeking proximity to the caregiver is especially apparent from 12 months when infants are suddenly more mobile as they are able to crawl and then walk, (3) a child’s caregiver is a secure base from which they can explore, and (4) the caregiver is a safe haven to return to if the child encounters something distressing. 

All four of these components can also be found in dogs with respect to their owners. Dogs seek to be close to their owners, especially in stressful circumstances (proximity seeking). When the owner is not there, dogs can experience separation-related distress. And experiments that borrow methods from child psychology have demonstrated both the safe-haven and secure-base effects. 

A classic test of attachment between an infant and their caregiver, called the Strange Situation, is a standardized procedure in which the infant is in a room with their caregiver and then a stranger comes in. Initial attempts to replicate the Strange Situation with dogs had mixed results, probably because a well-socialized dog is happy to see a friendly stranger. So scientists designed a version that included a threatening approach from a stranger.

The results showed some similarity to the behavior of infants during the Strange Situation test; the presence of the owner had a “secure base” effect. During the threatening approach, the increase in dogs’ heart rate was not as great if the owner was with them. In the reactive dogs, if they first met the stranger when the owner was present, they were less stressed when subsequently meeting the stranger without the owner (though still not at baseline). The study also showed individual differences. Some dogs were interested in the stranger despite their threatening approach, whereas others reacted by growling and barking. 

But do dogs always prefer their owners? Dr. Erica Feuerbacher and Prof. Clive Wynne tested this idea by giving dogs a free choice for ten minutes. The dog’s owner and a stranger sat down and would pet the dog if the dog came to them. About 80 percent of the time, the dogs chose to be near a person, but which person depended on the setting. In a familiar setting, the dogs spent twice as much time with the stranger, and in an unfamiliar setting they spent about four times as long with the owner. Even in the home setting, most of the dogs went to their owner first before they went to the stranger. (Interestingly, when shelter dogs were tested with two strangers, they typically showed a preference for one of the strangers, showing that dogs can establish preferences for one person over another quite quickly.) This study showed that while dogs are keen to meet other people, they have a special bond with their owner. 

“In a familiar context the dog would sort of say hi, check in with the owner and then go and spend their time with the stranger,” said Dr. Feuerbacher. “Whereas in an unfamiliar setting, the dog was really reluctant. I think some of our dogs never went and even said hi to the stranger, they just hung out with Mom in that probably stress-inducing context.” 

There are important implications for how owners can help their dogs in a stressful situation. Feuerbacher suggested calm petting and being there for the dog would be beneficial.  

Canine Neuroscience and Dogs’ Preferences for People and Food 

Another strand of research that speaks to the human–animal bond comes from canine neuroscience. Prof. Gregory Berns and his team at Emory University in Atlanta used positive reinforcement to train dogs to voluntarily go into and keep still in an fMRI scanner. This research looked at a part of the dog’s brain called the ventral caudate, which, in line with earlier work on both monkeys and people, is activated in response to the anticipation of a reward. 

One study looked at activation of the caudate when the dog was exposed to different smells: a familiar human (the dog’s main caregiver), an unfamiliar human, a familiar dog, an unfamiliar dog, and the dog’s own scent. Twelve dogs took part, and the smells came from swabs taken from the armpit of humans and from the perineal-genital area of dogs. The caudate was activated significantly more in response to the smell of a familiar human than in response to any of the other smells, including the familiar dog. So dogs recognized the smell of their main caregiver and had a positive association with it. While it’s hard to interpret these results in terms of the dog’s subjective experience, it shows the importance of the person to their dog. In humans, the same part of the brain is activated in response to looking at photos of loved ones. It’s unlikely that this is just a conditioned response to the caregiver being the one who feeds the dog. 

Another study from Berns’s lab investigated dogs’ preferences for food, the sight of the handler who praised them, and a control condition in which nothing happened. Since the dog has to keep absolutely still in the scanner, scientists did the experiment by pairing a different item (a toy car, a toy horse, or a hair brush) with each of the three conditions. Each item was on a stick and was presented to the dog for ten seconds, and then the relevant event happened. At a general level, there was no significant difference in activation of the caudate between the food condition and the praise-plus-sight-of-human condition, suggesting dogs found both rewarding. But there were individual differences between dogs: nine had roughly equal responses to food and praise, four preferred praise-plus-sight-of-human, and four preferred food only. 

Berns told me in an email, “The takeaway is that dogs, like people, are individuals, and that there is a spectrum of motivations. Some prefer food, some prefer praise, and many like both equally. Know which your dog prefers!” 

Do Dogs Understand Human Emotions? 

There’s evidence dogs respond to human emotional expression too. In one study, pet dogs were tested in their home to see how they responded to their owner or a stranger pretending to cry or humming (a control condition). Dogs paid significantly more attention when their owner was crying rather than humming. If the crying had made the dogs feel sad, you would expect them to go to their owners to be comforted. But the dogs directed their behavior to the person who was crying, regardless of who it was. Most of the dogs actually approached the person who was crying, and almost all did so in a way that could be seen as trying to comfort them. This doesn’t necessarily mean the dogs had empathy, but they approached the crying person in a way that the person would interpret as trying to comfort them. 

Another study had the dog’s owner sitting behind a see-through door and either crying or humming. In both conditions, dogs were equally likely to break through the door, but they were faster if the person was crying. Dogs who opened the door in response to crying had lower stress levels than those who did not; these dogs paced, fidgeted, and barked. As well, this study found dogs who were more attached to their owner were more likely to open the door in response to crying, which suggests the possibility they were feeling empathy. 

How People and Dogs Interact 

Dogs have their own ways of asking for things, like when they want tummy rubs. A study published in Animal Cognition analyzed 242 episodes of communication in videos of thirty-seven dogs interacting with their owners. This was a citizen science project in which people took video at home and sent it in to the researchers. 

The scientists found nineteen different gestures that, at least a lot of the time, satisfied the five criteria for referential gestures: (1) it is a gesture (rather than a movement that physically does something), (2) it is directed at an object or a part of their body, (3) it is aimed at someone else, (4) it gets a response from them, and (5) it is intentional in nature. Several things might show it is intentional, including waiting for a response and repeating the gesture if it has not worked to get what they want, or trying other gestures to get the outcome. In dogs, all five of those criteria were met. 

The dogs’ gestures were used to mean four different things: “scratch me,” “give me food/drink,” “open the door,” and “get my toy/bone.” Some gestures, such as pressing the nose or face against a person or object, could be used for all four meanings, while others were only used for some. For example, rolling over always meant “scratch me.” The head turn, which involves looking from an object to the human and back, was the most common gesture. What is especially impressive about these results is that they apply to communication with another species (us)! 

Play is an important part of interactions between people and their dogs. Even when dogs have the opportunity to play with other dogs, they still like to play with people. When playing with humans, dogs are more likely to show toys and present them to their play partner, suggesting that play with people has a different motivation than play with another dog. Tug, chase, chasing objects, and showing objects all occur more often in play between a dog and a person than between two dogs. 

People use a range of gestures to encourage dogs to play with them. One study looked at video of twenty-one people initiating play with their dogs at home without using toys, and followed it up by testing signals on twenty Labrador Retrievers. In a surprising result, although some actions were better at encouraging dogs to play, they were not the actions used the most often. The most effective signals were when the human bobbed like a play bow, chased or ran away from the dog, or lunged at the dog. Tapping the chest or otherwise signaling the dog to jump up also often worked, but was not used very much. Some of the techniques people tried, including grabbing the dog by the scruff, stamping, and picking up the dog, never worked to induce play. 

“I think the world would be better for dogs if people stopped to consider things (all the things!) from the canine point of view. Dogs in companion and working roles are often put into situations they wouldn’t choose for themselves. We can improve dogs’ quality of life by considering our decisions that impact them—how long we leave them alone each day, where they live, what training techniques we use, how we transport them, what we expect them to tolerate (from interactions with children, to dress ups and involvement in other human pursuits—like sky diving) and ask ourselves at every step, ‘Is this what my dog wants to do, if given a choice?’ Not all situations where our dogs would choose differently are avoidable (e.g., temperature taking at the vet clinic) but people should consider dogs and the way their lives are lived from the canine perspective. To see dogs as individuals who can experience a range of emotions with the capacity to suffer or thrive, both physically and mentally, based on the decisions we make about their lives—rather than just assuming that dogs like what we like, or that they are there to meet our whims, provide us with utility benefit, or be our entertainment—would be a huge advance for many dogs in our world.” 

— Mia Cobb, PhD candidate, Monash University, and co-author (with Julie Hecht) of the Do You Believe in Dog? blog 

We are responsible for providing everything our dogs need, and the way we do so can make a difference to how happy they are—or whether they’re happy. 

How to Apply the Science at Home 

• Understand that you are important to your dog. Your presence can give your dog the confidence to explore new things, and your dog will look to you for information when presented with a new or stressful item.

• Know that your dog can tell whether you are happy or sad, and this may be why dogs seem to comfort people when they need it.

• Although there is no specific research on how long a dog can be left on its own, and it will depend on the dog, use four hours as your maximum guideline. If your dog has to be routinely left for longer than that, consider asking a friend or neighbor to pop in and see the dog, arrange for a dog walker, or find a good doggy daycare. 

Adapted with permission of the publisher from the book Wag: The Science of Making Your Dog Happy written by Zazie Todd and published by Greystone Books in March, 2020.