The Dog’s Mind

An interview with Brian Hare, PhD, a leading researcher into canine cognition.


In the not-too-distant past, if you heard the word “dog” and “mind” in the same sentence, someone was probably talking about obedience, as in: “My dog minds pretty well.” Or, “You’d better mind me, or else!” Today, if you hear those two words in relation to each other, you are at least as likely to be listening to someone talking about canine cognition – the fascinating possibility that dogs are far more able to process thoughts and grasp concepts than previously given credit for.

Only in the past 15 years has the domestic dog begun to be accepted as a study subject for behavioral research.  Brian Hare, PhD, professor of evolutionary anthropology at Duke University, is one of the people who have legitimized the field, and I was thrilled to have the opportunity to interview him recently, on the occasion of the publication of his enlightening new book, The Genius of Dogs: How Dogs are Smarter Than You Think (Dutton, 2013).

Dr. Hare opened the Duke Canine Cognition Center (DCCC) in the fall of 2009. With his wife and co-author Vanessa Woods (a research scientist at DCCC, as well as an award-winning journalist and author of Bonobo Handshake), Hare wrote the book to provide a comprehensive review of what they’ve been studying at the DCCC – everything about dog cognition or, as they call it, “dognition.” Their goal was to bring historical and current information about canine cognition to the general public.

The Genius of Dogs covers a lot of material. Dr. Hare writes about his own dog-related experiences, professional and personal, such as visiting Russian scientist Dmitrii Konstantinovich Belyaev’s famous silver fox breeding compound in Novosibirsk, Siberia (Belyaev died in 1985, but others continue his work there); working with New Guinea Singing Dogs at the New Guinea Singing Dog Conservation Society in Eugene, Oregon; and struggling with training challenges he encountered with his own dog, Milo, a probable Lab/Chow mix. It’s a fascinating book, even given that other behavioral scientists have been critical of some of Hare’s conclusions about canine cognition, accusing him of overreaching at least a little.

Hare was a delight to interview. He is wildly enthusiastic about his work and can only be described as effervescent when speaking about his book, the Dognition website, and canine cognition in general.

Pat Miller: Your new book on canine cognition, The Genius of Dogs, co-written with your wife, Vanessa Woods, recently made the New York Times Best Seller list. Why did you write it?

Dr. HARE: There were a few things we wanted to communicate:

1) Science is really excited about dogs. That’s not always been the case.

2) There was not a book written that tried to do a historical review of the field of behavior aimed at the general public/dog-owning audience. We wanted to write that book and make the information accessible.

3) We wanted to explain how scientists think about intelligence.

4) And we wanted to discuss how an evolutionary approach can help us understand our dogs – and ourselves.

Miller: Thirty years ago, the words “canine” and “cognition” weren’t often used in the same sentence. Why has that changed?

Dr. HARE: Because a cognitive revolution has occurred, just in the last 10 to 15 years. What’s going on in the minds of animals has become central to our understanding of psychology.

Miller: You say that another cognitive limitation is that dogs do not understand what someone knows or doesn’t know. How do we know that? Is it possible we may find out otherwise down the road, just as we once said animals didn’t feel pain, couldn’t use tools, didn’t have emotions?

Dr. HARE: It’s absolutely possible. It’s important to understand what science is: It is fun and powerful and you can be part of it (you don’t have to have a PhD). It’s also fluid. In part it’s about discovering “truth” – and even more, it’s about falsification of past conclusions. The entire thesis of this book could be wrong – and could be falsified by future studies. Science is eternally a work in progress.

Miller: How is “understanding what an owner can see” (p. 245) different from “understanding what someone knows or doesn’t know”? Is it about what the person knows in the present versus what the person knew in the past?

Dr. HARE: That’s a great question, and one that can keep cognition scientists engaged in long discussions. “Understanding what an owner can see” is, in a way, understanding the geometry of the situation – understanding what the human knows in the present. The dog can see how the human orients in relation to what the dog does.

“Knowledge” is much more complicated – it involves an understanding of what was known in the past. We don’t currently think dogs can do this. We could be wrong.

Miller: You’re kind of hard on trainers and behaviorism in the last part of your book. You relegate trainers to two categories, neither of which is very flattering: the “top dog” school and the “more is better” school. Is there not a third group that currently exists – trainers who meld the work of Pavlov and Skinner with an interest in and awareness of dogs’ cognitive abilities? What role does behaviorism play, if any, in your vision of the perfect approach to dog training?

Dr. HARE: It was not my intent to be harsh toward trainers, and I apologize if that’s how we came across. I have great respect for trainers; I’m a pitiful trainer!

The goal was to say “Let’s look at these two schools of thought, and let’s look at what’s in the literature about them. And in fact, there’s not much about dog training in the scientific literature. It’s an opportunity that begs for attention.

I actually do think trainers are using a cognitive approach – because they use terms like “the dog knows,” and “the dog wants to perform.” Those are terms of cognition, not behaviorism. But trainers haven’t had access to the literature. In academia, behaviorism says operant and classical conditioning are the only approach to canine learning. There’s no room for cognition in behaviorism. And it’s not that operant and classical conditioning don’t work – of course they do. It’s just that they are one kind of intelligence and learning – one kind of many occurring in the dog’s mind. I think trainers know this. It’s perhaps just a miscommunication about definitions.

Miller: What’s your opinion of what you call the “top dog” or “alpha” approach to dog training?

Dr. HARE: Again, I’m not a trainer, but let’s look at the science. Is the rationale behind the alpha approach to training backed up by good science? There is not good evidence in the literature, and it doesn’t make much sense. In fact, the alpha thing is based on a big mistake: using wolves as a model for dog behavior.

It’s easy to get befuddled by evolution. Dogs are descended from wolves, therefore dogs are like wolves. In some ways, this is correct. But it’s also correct that dogs are a different species than wolves, and therefore they are not like wolves.

In fact, a much better model for dog behavior is the behavior of thriving packs of feral dogs. The feral dog social system is nothing like that of wolves: there is no alpha pair, no reproductive suppression, no infanticide, no fatal aggression over territory. In fact there is very little aggression in feral dog packs; the leader is simply the dog with the greatest number of affiliative relationships – the dog with the most friends.

There is one caveat – a significant difference between feral dogs and our dogs. In an established feral pack, over generations the dogs all end up being of similar size (a medium size) at maturity. The similarity in size helps to minimize aggression. We have such a huge variation in size in our companion dogs, so we do see significant aggression, both in our homes as well as places like dog parks.

Miller: You talk about the dog’s ability to “learn how to learn” as a part of cognitive training. How does this differ from what modern trainers call “learning how to learn” as a part of positive reinforcement-based training?

Dr. HARE: It probably doesn’t. It’s just an acknowledgement – one that many trainers seem to be aware of and are making use of – that canine learning goes beyond the behaviorism interpretation of simply learning and solving the same problem over and over again. Rather, dogs, through learning and experience, possess a learning set; they can actually form a concept and apply it to new problems. That’s cognition. The idea would make Skinner roll over in his grave.

Miller: You say (on page 246) “When an experimenter shows where food has been hidden but then points at another location, dogs do not search for the food they saw but instead go to where the human pointed.” I did the Dognition games yesterday with my Corgi, Lucy, and she went for the food she had seen 100 percent of the time, rather than where I pointed. Can you explain?

Dr. HARE: Studies involving dogs utilize a small sample size – usually fewer than 30 dogs. We develop group statistics based on this sample and then try to apply our conclusions to the whole population. It doesn’t necessarily mean they all do it. So if, in our study, 60 percent of the dogs go to where the human pointed, we say that’s what dogs generally do, even though 40 percent of the dogs did something else. Some dogs, like your Lucy, rely on their own memory (knowledge) more than they rely on signals from their humans.

Miller: You’ve also recently launched the online service “Dognition.” Can you describe what is offered at Dognition? Why did you create it?

Dr. HARE: As a scientist, I’m excited about outreach – communicating to people that science is fun and powerful, and they can participate in it. I see it as a service: to give people a fantastic experience and the opportunity to gain more information and better explanations about their dogs’ behaviors, to help them see and understand things about their dogs, and to enhance their relationships with their dogs.
Of course, we’re seeking to make discoveries about dogs, and to have fellow dog lovers participate in making those discoveries. We are also working to be good canine citizens; we’re offering Dognition memberships to shelters for free, to give them another tool for working with the dogs in their care.

We also plan to fund more behavioral research, and look forward to working with trainers to develop an even more cognitive approach to dog training. We really would like to see the application of our research – to help trainers see things about dogs they wouldn’t see otherwise. Our ultimate goal is to see a quantum leap forward in our ability to help dogs.

We’re already happy with what we’re seeing on the site. It’s very rewarding to see the many comments like, “I had no idea my dog was capable of doing that!”

Subscribers to Dognition will receive a new game every month to play with their dogs. They will be the first to see new discoveries in our work with canine cognition, and have access to scientists’ comments and scientific content.

The punch line for Dognition – it’s a place where you can play fun games with your dogs. “Just like we feed their stomachs, we need to feed their minds.” Dognition can help people understand that.


We tend to think of intelligence as that which can be measured by an IQ test. By that standard, our dogs fall low on the scale. But there are many different ways to measure a dog’s capabilities. If dogs were to develop an IQ test it might well rely on the ability to find scent – in which case they would be brilliant and we would be the dullards!

Inferences:  Having heard about Rico, the German Border Collie who learned the names of more than 200 objects, retired psychology professor John Pilley adopted an 8-week-old Border Collie pup and decided to see how many object names his dog could learn. Over a period of three years, Chaser learned the names of more than 1,000 different objects. Both Rico and Chaser, when asked to fetch an object they weren’t familiar with, unerringly picked the one object in the group for which they hadn’t learned the name. The dogs inferred that it must be the correct object, since they knew the names of all the others. Researchers tell us that this is similar to how children can learn the names of new objects.

Pilley took this research game with his dog one step further. Children are able to categorize objects. “Sock,” for example, is not just one object that happens to be a sock, but rather is a category name for all sock-objects of different sizes, colors, shapes, and textures. When scientists suggested that babies could learn words as categories and dogs could not, Pilley took up the challenge. He taught Chaser that his toys were classified into different categories. Then he mixed toys from different categories, and asked Chaser to fetch a category of toy. Chaser performed flawlessly.

Symbols: Canine cognition doubters also suggested that if dogs truly learned words, they should also be able to learn symbols. If you show a child a replica of a toy, a child understands it is a representation of a real life thing. If you show a child the replica and ask him to go get the thing, he can do it. Could a dog? Dr. Juliane Kaminski, the scientist who did the original studies with Rico in 2004, pursued this question. She used Rico and several other Border Collies, asking them to fetch toys in another room by showing them a replica of the toy, rather than using the name of the toy. Some of the replicas were the same size, some were miniatures. All of the dogs were successful at retrieving the correct object after being shown the replica. Rico and one other dog were even able to retrieve the correct toy when they were just shown a photograph of the object.

This means dogs are able to grasp concepts – the idea that something can symbolize something else. We can only wonder – and wait – to see what other incredible things the world of canine cognition science has yet to uncover about our dogs’ intelligence. Or maybe you don’t have to sit idly by and wait. You and your dog can participate!

Dr. Hare is not the first nor the only researcher working to bring attention to the cognitive abilities of the canine mind. Even the facility he founded, the Duke (University) Canine Cognition Center is not unique; similar facilities now operate at Harvard, Barnard College, the University of Florida, and many other places; in fact, you can find them all across the U.S. and Europe.

In her 2010 book Inside of a Dog: What Dogs See, Smell and Know, cognitive scientist Dr. Alexandra Horowitz, who runs the canine cognition lab at Barnard College, addresses the workings of the canine brain.

Dr. Marc Beckoff, professor emeritus of ecology and evolutionary biology at the University of Colorado, Boulder, touched on the subject back in 2002 in his book Minding Animals; Awareness, Emotions and Heart (this book is mostly about wild animals), and has written extensively about canine ethics, emotion, and intelligence over the past decade. In a July 19, 2010 post on his “Animal Emotions” blog on, Beckoff ascribes to dogs the quality of metacognition: the ability to know what someone else knows, a concept that was long held to be a prerogative of the human brain.

Dr. Marc Hauser, an evolutionary biologist and a researcher in primate behavior, animal cognition, and human behavior, was a Harvard University professor from 1992 to 2011, regarded as “a leader in the field of animal and human cognition”. In 2011, the university found him guilty of scientific misconduct (fabricating and manipulating research data in some monkey studies) and he resigned. Reportedly, he now works with at-risk youth.

As part of his ongoing work on canine cognition, Hare and his colleagues have created a website ( that invites – nay, begs for! – public participation. Using suggestions on, you and your dog have fun playing cognition games together. You learn more about your dog, and your results are compiled as data for current and future studies.

According to the website, “A key aspect of the Dognition methodology is our use of Citizen Science – research that can be conducted by everyone, not just people with PhDs. By gathering this data we can begin to understand more about all dogs, much more quickly and on a broader scale than if scientists had to conduct this research themselves.” Citizen science. I love it!

In preparation for my interview with Dr. Hare, I worked through the Dognition Toolkit games with my 9-year-old Cardigan Welsh Corgi, Lucy. It was fun – and a little stressful (for example, when I had to set her up to “disobey” her “Leave it” cue). I was bemused to discover in one of the games that Lucy does not do what the majority of dogs do: Rather than going to the empty upside-down cup I point to, she consistently goes to the upside-down cup on the opposite side, where she saw me hide the treat . . . meaning she relies on her own observation more than she relies on me indicating where the treat might be. Isn’t that just like a herding dog?

I recently paid the membership fee to join Dognition for a year. I admit I initially balked at the idea of paying to play, but I really am enchanted by the idea of citizen science, and the curious part of me couldn’t resist the opportunity to see some of the results. And hey, Lucy and I get to be a part of making cognition history. How could we resist that? Maybe we’ll see you there.

Previous articleDoxycycline Shortage leaves People and Pets Scrambling for Treatment Alternatives
Next articleEspecially Embarrassing When Your Good Dog Is Bad
Pat Miller, CBCC-KA, CPDT-KA, grew up in a family that was blessed with lots of animal companions: dogs, cats, horses, rabbits, goats, and more, and has maintained that model ever since. She spent the first 20 years of her professional life working at the Marin Humane Society in Marin County, California, for most of that time as a humane officer and director of operations. She continually studied the art and science of dog training and behavior during that time, and in 1996, left MHS to start her own training and behavior business, Peaceable Paws. Pat has earned a number of titles from various training organizations, including Certified Behavior Consultant Canine-Knowledge Assessed (CBCC-KA) and Certified Professional Dog Trainer - Knowledge Assessed (CPDT-KA). She also founded Peaceable Paws Academies for teaching and credentialing dog training and behavior professionals, who can earn "Pat Miller Certified Trainer" certifications. She and her husband Paul and an ever-changing number of dogs, horses, and other animal companions live on their 80-acre farm in Fairplay, Maryland.