A shaggy brown terrier approaches a large chocolate Labrador in a city park. When the terrier gets close, he adopts a yogalike pose, crouching on his forepaws and hiking his butt into the air.
The Lab gives an excited bark, and soon the two dogs are somersaulting and tugging on each other's ears. Then the terrier takes off and the Lab gives chase, his tail wagging wildly. When the two meet once more, the whole thing begins again.
Watch a couple of dogs play, and you'll probably see seemingly random gestures, lots of frenetic activity and a whole lot of energy being expended. But decades of research suggest that beneath this apparently frivolous fun lies a hidden language of honesty and deceit, empathy and perhaps even a humanlike morality.
Take those two dogs. That yogalike pose is known as a "play bow," and in the language of play it's one of the most commonly used phrases. It's an instigation and a clarification, a warning and an apology. Dogs often adopt this stance as an invitation to play right before they lunge at another dog; they also bow before they nip ("I'm going to bite you, but I'm just fooling around") or after some particularly aggressive roughhousing ("Sorry I knocked you over; I didn't mean it.")
All of this suggests that dogs have a kind of moral code -- one long hidden to humans until a cognitive ethologist named Marc Bekoff began to crack it.
A wiry 68-year-old with reddish-gray hair tied back in a long ponytail, Mr. Bekoff is a professor emeritus at the University of Colorado at Boulder, where he taught for 32 years. He began studying animal behavior in the early 1970s, spending four years videotaping groups of dogs, wolves and coyotes in large enclosures and slowly playing back the tapes, jotting down every nip, yip and lick. "Twenty minutes of film could take a week to analyze," he says.
The data revealed insights into how the animals maintained their tight social bonds -- by grooming each other, for example. But what changed Mr. Bekoff's life was watching them play. The wolves would chase each other, run, jump and roll over for seemingly no other reason than to have fun.
Few people had studied animal play, but Mr. Bekoff was intrigued. "Play is a major expenditure of energy, and it can be dangerous," he says. "You can twist a shoulder or break a leg, and it can increase your chances of being preyed upon. So why do they do it? It has to feel good."
Suddenly, Mr. Bekoff wasn't interested just in behavior; he was interested also in emotions and, fundamentally, what was going on inside these animals' heads.
Mr. Bekoff wasn't the first scientist to become intrigued by the canine mind. Charles Darwin in the mid-1800s had postulated that canines were capable of abstract thought, morality and even language. (Darwin was inspired by his own mutts; he owned 13 of them during his life.) Dogs, he wrote, understand human words and respond with barks of eagerness, joy and despair. If that wasn't communication between the species, what was?
Two of Darwin's contemporaries had suggested that dogs could even sniff out someone's social status and read words. But by the time Mr. Bekoff turned his attention to canines, scientists had long deemed them unworthy of study. Because they no longer lived in their natural environment, the thinking went, their minds were corrupted and could not shed light on the bigger question, the evolution of human intelligence. The only animals worth studying were wild ones.
But when Mr. Bekoff began looking at videos of dogs romping in super-slow motion, he began to realize that there was more going on in the canine mind than science had acknowledged. He noticed the "play bow," for example.
What's more, he found that canines "role-reverse" or "self-handicap" during play. When a big dog played with a smaller one, for example, the big dog often rolled on her back to give the smaller dog an advantage, and she allowed the other dog to jump on her far more often than she jumped on him.
Mr. Bekoff also spotted a number of other blink-and-you'd-miss-them behaviors, such as a sudden shift in the eyes -- a squint that can mean "you're playing too rough" -- and a particular wag of the tail that says, "I'm open to be approached." Humping a playmate during a romp, meanwhile, was often an invitation to nearby dogs to come join the fun.
Such signals are important during play; without them, a giddy tussle can quickly turn into a vicious fight.
In the wild, coyotes ostracize pack members that don't play by the rules. Something similar happens in dog parks: If three dogs are playing, and one bites or tackles too hard, the other two are likely to give him the cold shoulder and stop playing with him, Mr. Bekoff says. Such behavior, he says, suggests that dogs are capable of morality, a mindset once thought to be uniquely human.
Even morality hints at something deeper, however. To enforce moral conduct, dogs must be able to experience a spectrum of emotions, from joy to indignation, guilt to jealousy. They must also be able to read these emotions in others, distinguishing accident from intent, honesty from deceit. And indeed, recent studies by other scientists have shown evidence of these abilities (confirming what many dog owners already feel about their pets).
Scientists have found, for example, that dogs trained to shake hands with humans will stop shaking if they notice that they aren't being rewarded for the trick although a nearby dog is -- a sign, the researchers suggested, that dogs can sense inequity.
Other studies have revealed that dogs yawn when they see humans yawning, and that they nuzzle and lick people who are crying; scientists consider both behaviors displays of empathy, a rarely documented trait in the animal kingdom.
Dogs have even been shown to be pessimistic: When a group of canines in one study learned that a bowl placed on one side of the room contained a treat and a bowl on the other side contained nothing, some of the dogs just sat there when the empty bowl was placed in the center of the room; they figured it was empty and didn't waste their time. These same dogs evinced what researchers said was a similar pessimistic attitude when their masters left for work: They were more likely to howl and tear up the couch when their owner disappeared, possibly because they didn't believe that their master would return.
Mr. Bekoff's recent work suggests another remarkable canine skill: the ability to know what another animal is thinking -- a so-called "theory of mind." Dogs seem to display a rudimentary form of this skill during play. He has noticed, for example, that one dog won't begin trying to play with another dog until he has her attention. To get her to notice, he may nip the other dog or run into her field of view. That, Mr. Bekoff says, shows that the one wanting to play knows that she's not paying attention to him.
Though this may seem like a simple skill, it's incredibly important to our species. Without it, we can have a hard time learning or interacting with the world around us.
Interestingly, dogs even outsmart chimpanzees on some theory-of-mind tests. When a researcher points at one of two cups, for example, dogs almost always run to the cup that is pointed to, a sign that they have intuited what the scientist was thinking -- i.e., that the researcher was trying to show the dog something. Chimps, by contrast, have no idea what we mean when we point at something.
"Dogs have an amazing relationship with us, and Marc [Bekoff] has done a beautiful job helping us understand them," says Duke University biological anthropologist Brian Hare, one of the world's foremost experts on canine cognition. "Play gives us a peek inside their heads and helps us understand how they became the species they are today."
Mr. Hare, one of the first scientists to show that dogs could understand human pointing, while chimpanzees could not, says Mr. Bekoff's studies add a new dimension to the canine personality: Dogs aren't just smart, they're also emotionally complex. "That's why we can have such a deep relationship with them," Mr. Hare says.
It's also why studying dog play is so important, Mr. Bekoff says. It reveals far more than just the emotional lives of the animals involved. It could ultimately shed light on the evolution of human emotions, and how we came to build a civilization based on laws and cooperation, empathy and altruism.
Play may seem a frivolous activity, but because it is not simply a survival reflex, it provides the best opportunity to explore who the animal really is, to peer perhaps into her soul. "When we study play in dogs," Mr. Bekoff says, "we study ourselves."
David Grimm is a deputy news editor at Science and the author of the new book "Citizen Canine: Our Evolving Relationship With Cats and Dogs."