Belyaev and his friendly foxes

You pretend to throw a Frisbee in one direction, then toss it in the other. Your dog swiftly shifts gears to catch the disk in mid-air. Or you’re in a bad mood and your pup intuits that a sloppy kiss will cheer you right up.

All dog owners have experienced, and likely marveled at, such examples of their pets’ observation skills. But it shouldn’t surprise us that dogs are adept at gauging our behavior. Nicholas H. Dodman, director of the Animal Behavior Clinic at Cummings School of Veterinary Medicine at Tufts University, writes in The Well-Adjusted Dog “Thousands of years of cohabitation of dogs with humans have honed dogs’ inborn skills to the point where, even as pups, they can understand what we mean by certain gestures.”

One gesture in particular has attracted the attention of scientists recently: Pointing.  According to Brian Hare, assistant professor of evolutionary anthropology at Duke University and director of the Duke Canine Cognition Center, the ability to track the direction of our arms, fingers, and even eye movements places canine social intelligence at approximately the level of that of a 2½ year old human — and beyond the capabilities of dogs’ likely ancestors, wolves, as well as those of our primate cousins, chimpanzees and great apes. Dogs will look towards the direction where people point — that’s why it’s fun to try to fool them with Frisbee feints — whereas wolves, even those reared by humans from birth, will just look at the hand itself.

The discovery that even puppies have the skill to follow pointing, while wolves raised in captivity by humans do not, suggests that it’s a trait passed along genetically. Stanley Coren, Professor of Psychology at the University of British Columbia and author of several books about canine cognition, says, “We have been systematically and selectively breeding dogs so they read our body language. If you have two animals and one responds well to your gestures, that one is going to get preferred treatment and will be selected for breeding.”

Indeed, this process can occur rather quickly if nature gets a nudge. In the 1950s, Soviet geneticist Dmitry Belyaev conducted a landmark experiment with silver foxes in Novosibirsk, Siberia. His only criterion for choosing specimens to breed was tameness; the foxes he selected were less afraid of humans than the others. Within eight generations, Belyaev began to see foxes that not only behaved like dogs, but had canine physical characteristics such as floppy ears and curved tails.

Belyaev initially created his tame silver foxes to benefit the fur industry — the wild foxes were too stressed to breed successfully — but some 40 generations later, the animals are attracting scientific attention for their own sake. In 2003, Brian Hare traveled to Novosibirsk to test whether baby foxes that had been bred to be doglike responded to his pointing test the same way that puppies did. Sure enough, Hare discovered, they had the same ability to respond to human signals as domesticated canines.

While it’s clear why humans would breed dogs to do our bidding, we may wonder what’s in it for dogs, genetically speaking. Do they only want food and shelter from us?

No, asserts Dr. Dodman, although selfishness is a common characteristic of all animals, including humans. With dogs, he says, “You have very friendly group dwellers who benefit from a relationship with people. And it’s not just because we feed them.” In his behavioral clinic, Dr. Dodman sees plenty of evidence that dogs enjoy our physical presence. “When we leave them, something like 15 to 17% of American dogs have been described as having separation anxiety,” he says. “They’re not anxious because they don’t know where their next meal is coming from. They’re anxious because they’re separated from their family, their humans.”

Dogs are also interested in observing us because they are neotenized, according to Dr. Dodman. “Dogs are unique in that they are pack animals that have been bred to look and act childish. They are permanently stuck in canine Peter Pan mode,” he explains. Not only are their physical characteristics similar to those of the young in other canids, but  their behavior is also childlike. They are “friendly, dependent, playful, and smiley, and they follow our directions,” Dr. Dodman says. And like children, he adds, “They have a desire to follow the leader, to be like the adults — or in this case, the humans.”

But social intelligence, which sets dogs apart as a species that interacts uniquely with humans, is not the only type of brainpower by which we judge dogs. In Intelligence of Dogs, Coren also discusses “adaptive intelligence,” which measures what dogs can learn to do for themselves and how much they can benefit from interactions with their environment. Unlike social intelligence, adaptive intelligence can differ within a breed.

“Dogs are constantly learning,” Dr. Dodman says. They remember kindnesses and unpleasantness their whole lives, and learn to avoid things that are bad for them.” That includes people, in many cases. Assuming that humans can’t be trusted is not a bad thing, according to Dr. Dodman, unless it’s taken too far.  “A dog may take out a class action suit against men with beards,” he says. Furthermore, they may react to things they are averse to in a fashion that humans find socially unacceptable. “Dogs discover that the best way to make something go away is to bark and lunge at it,” Dr. Dodman says. “It works every time.”

Coren calls the third type of canine intelligence “working and obedience intelligence.” It includes a dog’s ability to learn words or commands — a  smart dog can learn as many as 150 of them — which serves them well in playtime activities that involve complex routines such as agility; and in the jobs they tend to be offered today (as opposed to sheep herding): leading the blind, say, or conducting search-and-rescue missions. “It’s the closest to what we might call school-learning ability,” Coren observes. “And it is based on what the dog can learn to do when instructed by humans.”

But it is their instinctive intelligence, their ability to read us, that has caused the most confusion about what dogs know and how they know it. In what is sometimes called the “Clever Hans effect” after an early 20th-century German horse whose touted mathematical abilities were later discovered to stem from an ability to read his handler’s body language, dogs were long thought to be solely reactive to humans, and thus not capable of more sophisticated learning.

At the same time, their powers of observation have caused humans to erroneously attribute complex emotions to dogs. Owners often claim, for example, that their dogs feel guilt over their misdeeds. Alexandra Horowitz, a cognitive scientist at Barnard College and author of Inside of a Dog: What Dogs See, Smell, and Know, conducted experiments to determine if dogs who misbehaved and were scolded by their owners responded differently to tongue-lashings than dogs who didn’t misbehave but whose owners scolded them anyway (they had been told that, when they left the room, their dogs had been bad). The results were conclusive: It was the owners’ signs of anger that had caused the dogs to lower their tails and bow their heads, not any knowledge that they had actually committed an act for which they were accustomed to being punished.

No question. Dogs use their observation skills to modify human behavior. They have learned that, if they look remorseful, we will stop barking at them and treat them with affection again. Luckily, they don’t take their primate studies to a more abstract level. We love dogs because they play Frisbee with us without trying to figure out how our motor skills evolved and because they cheer us up without inquiring if we’ve ever considered Freudian therapy.


A version of this article originally appeared in Your Dog, the newsletter of Cummings School of Veterinary Medicine at Tufts University. I can’t recall why, but all the references to the pointing experiments, foxes, and Alexandra Horowitz were edited out. Maybe they were too esoteric? Forgive me if you agree. I’m a bit of a geek.

This post makes it clear that dogs have the ability to read us, and gives a sense of why they might want to. The second part of this article will detail some of the mechanisms behind dogs’ mind-reading abilities.

As for the Freud reference at the end: I must have channeled the future. Or maybe it was my resident mind reader, Frankie, who told me…


21 thoughts on “How Dogs Read Us, Part I”

  1. I’ve always been curious about the dog’s ability to follow where I point. Glad to hear an explanation. Although I’m distrustful of studies. You’ve mentioned the “guilty” study before, and I swear when I had my very smart dog Pumpkin, I would return to the house and before I had said a word or done anything except walk in the door, she would be hanging her head if she had emptied a wastebasket. Maybe she wanted to be sure that I KNEW she emptied the wastebasket?

    1. Or maybe the brilliant Pumpkin was the exception that proved the rule? I’m glad I could offer an explanation of the pointing-follow ability, your distrust of studies notwithstanding.

  2. I hope the Tufts newsletter was just editing for space because I found the descriptions of the research very concise and readable. I loooooooove writers who can make boring old science interesting to read.

    You’ve just joined my list alongside Diane Ackerman and Mary Roach. 🙂

    I also find it interesting that dogs are so much better at reading body language than humans are. I’ve spent many hours reading about and watching dogs. But I find they respond to bodily cues more quickly than I can keep up. It’s only after I see photos taken at the dog park, for instance, that I finally understand what caused a dog’s strange reaction to something.

    I suspect this is a case where our ability to speak limits us in other forms of communication.

    1. I’m honored to be mentioned in the same sentence as Mary Roach, one of my favorite writers! Thank you.

      I agree that our ability to speak — and to let our minds wander, aka daydreaming — helps account for our inability to keep up with our dogs when it comes to reading cues.

  3. I’ve read so much about dogs’ natural ability to follow human pointing but until very recently, my dog has never been able to do this. Even if I was pointing directly to a treat on the floor in front of me, she would keep sniffing everywhere but where I was pointing. Since she also seems to lack the ability to tell when another dog wants to play, I wonder if this pointing skill has something to do with early socialization, which I know my dog did not receive. Either that or just is just broken. 😉

    The silver fox experiment is endlessly fascinating. I never get bored reading about it.

    1. I like to think there is a perverseness factor that is little discussed — at least that’s my rationale for Frankie’s refusal of my pointing help to find the squeaky carrot…

    2. I’ll suggest one more issue with the pointing experiment. If you’ve ever seen it, a handler sits with the dog while the dog stays. The experimenter makes a pointing gesture to the cup hiding the food while the dog is watching the entire time.

      I find dogs respond better to motion than still signals. So if Shiva (or Frankie) aren’t watching you when your arm moves into a point, it might not mean much to them. I bet fewer dogs would pay attention to the pointing if the dog were not sitting at attention and already watching their person.

  4. I prefer to just think you are on-trend. Saturday night’s episode of Too Cute on Animal Planet had a trivia question about Freud and his dog!

  5. I knew these articles would be fun! Loved Dodman’s remark about dogs who take out a “class action suit” on men with beards – reminded me of Shepherd we adopted out. I look forward to your next installment.

    Editors – always a mystery. You may be right – the level of detail we like to have versus what they want to offer in a newsletter – puzzling considering who they are.

    1. Like God, editors have all the power and work in mysterious ways. Except when I am an editor, of course…

  6. It took Bella a while to recognize the pointing gesture, but like Kristine, I wonder if it was an issue of early socialization. Now she’s a champ at it though. I’m always kind of proud of her when she looks at what I’m pointing at.

    Very interesting post, Edie. I do think that dogs are probably better at picking up on those subconscious cues and subtle body language that we humans often overlook.

    1. Thanks, AJ. Yes, we are too often blissfully oblivious that we are sending out all kinds of signals for our dogs to read — often at odds with our verbal cues. What confusing creatures we are!

  7. As we humans have evolved so have other species. Dogs have always been kept as a pet. In the article as u have specified if wolves grow with humans they too become dependent. So when we consider dogs who have cohabited with humans it is but natural for them to read basic human emotions, because basic human emotions haven’t changed.

  8. Very cool.
    Yeah, I’m like you I like the fox info to be left in. Not the first I’ve heard of bad editor decisions though, Micheal Eades had the same problem with one of his when the publisher took out a chapter that most of us followers of him think was fantastic back story and old science.
    I also like that science seems to be backing up my view of the intelligence of dogs and I think we ought to keep trying to make law reflect it too. If you can’t abuse or kill a two year old, you ought not be able to abuse or kill a dog.

    1. Thanks, Jenni. And just to clarify — my editor is very nice and very smart and my stories are usually published more or less intact. This was a rare exception.

      Sadly, there are people who abuse and kill two-year olds too. But at least a change in laws would put animal abusers away for a longer time…

  9. Dogs may not feel guilt… but they will tell you when they’ve been up to something they shouldn’t have… I can tell by the way my dog acts when she’s been raiding the recycle bin.

    She’ll have ‘that look’ and I’ll start scanning the room and I’ll find a container every time. She’s very soft and I only have to speak with a disproving voice at her. And then it will be awhile before she gets tempted beyond her ability to resist…

    So there is a whole new set of experiments to set up.

Leave a Reply

Your email address will not be published. Required fields are marked *