Search the Site

Episode Transcript

America is a dog-loving nation. Hard numbers aren’t easy to come by, but it’s estimated there are between 77 and 90 million dogs in the U.S., with roughly 40 to 50 percent of households having at least one. We also have a lot of cats, but dogs appear to be a bit more popular. The pandemic has brought even more dogs into our homes, and made us even closer. So, how well do you know your dog? We may think we know them pretty well. But the fact that they are so familiar can actually make it harder to see dogs as they really are.

Alexandra HOROWITZ: Initially, studying dogs — having a familiar species as my chosen subject — was a little bit of a deficit because people felt like dogs were already understood, right? I mean, hopefully we know something about them because they’re living in our house and they’re in my bed right now, you know? And the thought was that you don’t really need research in this field. 

Alexandra Horowitz runs the Dog Cognition Lab at Barnard College in New York, which is part of Columbia University. She’s the author of several books about dogs, the best-known of which is called Inside of a Dog: What Dogs See, Smell, and Know. Here’s one brief passage that we asked her to read — page 7 of the paperback if you want to follow along.

I’ve gotten inside of the dog, and have glimpsed the dog’s point of view. You can do the same. If you have a dog in the room with you, what you see in that great, furry pile of dogness is about to change. 

Inside of a Dog was Horowitz’s first book. It was published in 2009 and it was a significant best-seller; now it’s considered a classic. At least I consider it a classic.

Go look at a dog. Go on, look — maybe at one lying near you right now, curled around his folded legs on a dog bed, or sprawled on his side on the tile floor, paws flitting through the pasture of a dream. Take a good look — and now forget everything you know about this or any dog. 

Today on Freakonomics Radio: we will take a good look at dogs, try to forget everything we know — and try to understand them from the inside out. It’s another installment of the Freakonomics Radio Book Club, in which we interview the author of a noteworthy book and hear some of the best passages read aloud. You’re going to like this one, I promise. Even if you are a cat person.

*      *      *

Alexandra Horowitz grew up in Colorado and always loved dogs, but it took her a long time before realizing she could study them for a living. As an undergrad, she was a philosophy major at Penn; she moved to New York and worked as a lexicographer for the Merriam-Webster dictionary and then as a fact-checker for The New Yorker. And there, she worked on some pieces by Oliver Sacks, the free-wheeling neurologist best known for writing Awakenings and The Man Who Mistook His Wife for a Hat.

HOROWITZ: And I just adored his combination of this empathetic science and sort of philosophical musings.  

At this point, she thought she might like to become that kind of writer. She went back to school and wound up with a Ph.D. in cognitive science from the University of California, San Diego.

HOROWITZ: It’s basically a Ph.D. that’s interdisciplinary. So, philosophers are asking, what is the mind? Neuroscientists are interested in examining the brain. Psychologists want to look at behavior. So, those fields plus computer science — with their interest in artificial intelligence — and anthropology, all kind of converge in this one interdisciplinary field, which wants to use different methods and approaches to tackle the questions of mind. 

Stephen DUBNER: And when you set out to get a Ph.D. in cognitive science, did you imagine that you would specialize in the cognitive science of not humans, but dogs? 

HOROWITZ: It was actually a surprise to me. But I was very interested in how you get answers about what animals who can’t give verbal responses are thinking. And dogs wound up being a good model for looking at behavior and making inferences to mind. So, it happened that way. But it hadn’t happened that way to people before, there wasn’t a cognitive science of dogs.

DUBNER: So, you were a dog pioneer, we should say. Yes? Don’t be shy. 

HOROWITZ: Well, right when I started looking at dog mind, there were other people around the world who also started looking at dog mind. So, there was this interesting convergence of thinking. So, there must have been something in the air, something smelly.

DUBNER: All of us are familiar with the study of cognition of nonhuman animals, especially different kinds of monkeys, chimpanzees, gorillas, and so on. But it strikes me that most of that is done as a kind of means toward an end, a better understanding of humans, at least to some degree. Whereas it seems as though once you got into dog cognition, you just really — I mean, you care about humans, plainly. But you really went whole dog, right?

HOROWITZ: I’m just interested in the dog as a dog, right? And because there is so much room to grow in the field, it’s not as though everything’s already been found out. So, there’s a lot of territory I could cover. 

Here, again, is Alexandra Horowitz reading from Inside of a Dog.

It perplexes me that some of the questions I have most often been asked about dogs, and that I have about my own dog, are not addressed by research. On matters of personality, personal experience, emotions, and simply what they think about, science is quiet. Still, the accumulation of data about dogs provides a good foothold from which to extrapolate and reach toward answers to those questions. The questions are typically of two kinds: What does the dog know? And what is it like to be a dog? 

*      *      *

DUBNER: You start the book writing about dogs at play. Why is that? Is play really so central to dogness?

HOROWITZ: One of the things I was super-interested in was metacognitive states — thinking about thinking — and theory of mind — thinking about myself and others’ minds as having different information than mine. We definitely learn a lot about other’s minds in social play. And so, I thought, “Well, why don’t we look at the behavior of social play of another animal and see if there’s some of this complexity of taking different roles, taking turns,” you know? 

And dogs just wound up being a really great subject to study because they’re playing all the time. And I could just go down to the dog park and suddenly I was researching my subjects. It almost seemed like a cheat. I mean, I will say it took me six months to come upon this realization that I should be studying dogs, even living with a dog and taking her to play a couple of times a day. But I finally did.

*      *      *

There are two ways to learn how play works — and what playing dogs are thinking, perceiving, and saying: be born as a dog, or spend a lot of time carefully observing dogs. The former was unavailable to me. Come along as I describe what I’ve learned by watching. In my study of dog play, I shadowed dogs with a video camera rolling, and controlled my own delighted laughter at their fun long enough to record bouts of play, from a few seconds to many minutes long. After a few hours of this the fun stopped, the dogs would get packed into the backs of cars, and I would walk home, reflecting on the day.

I’d sit down in front of my computer and play back the videos, at an extremely slow rate. Only at this speed could I really see what had happened in front of me. What I saw was not a repeat of the scene I’d witnessed at the park. At this speed I could see the mutual nods that preceded a chase. I saw the head-jockeying, open-mouth volleys that blurred into unrecognizability in real time. I could count how many bites it takes, over the course of two seconds, before a bitten dog responds; I could count how many seconds it takes for a paused bout to resume. I also noted their postures, their proximity to one another, and which way they were looking at every moment. Then, so deconstructed, the play could be reconstructed to see what behaviors match what postures. 

*      *      *

DUBNER: So, can you quickly describe, I guess, the status of dog research now versus 20 or 30 years ago?  

HOROWITZ: I mean, there were no research groups studying dogs. And now, there are dozens of dedicated dog cognition labs. I mean, it’s all over the world So, there’s a huge amount of research being produced. 

DUBNER: Now, not that you’re the type of person to say, “I told you so,” but if you were and you looked back at the beginning, at the peers or elders who discouraged or maybe even disparaged you for wanting to do serious cognitive-science research on dogs, who are so familiar and so on — if you were to want to rub it in a little bit now, what would you say would be the central findings or understandings of dogs that prove that they have been really worthy of serious study?

HOROWITZ: It’s not really like me to go back and rub their noses in it. And in fact, some of those people have come to study dogs themselves, so they already know the error of their ways. And I can just sit back satisfied in the knowledge that they had to have that realization. But I do think the dog’s amazing social cognitive skills, their ability to kind of solve puzzles using others, thinking about others’ minds, is one of the things that makes them extraordinary and something we didn’t expect to see except for in other primates. And we don’t even see it in all primates. So, that’s one thing. The other aspect is just the real interest in imagining a point of view which is not visually based.

DUBNER: So, let me ask you a quick SAT-style question. The sense of smell is to a dog as blank is to a human. 

HOROWITZ: Vision. They have this really dedicated instrument, their nose, the very outside of their nose is actually designed to catch a lot of odor molecules. It’s moist. That helps doing that. The area at the back of the nose, the olfactory epithelium, that has the cells that translate this odor molecule into experience of smelling something, you know, that area has hundreds of millions of cells in the dog, 300 million. 

DUBNER: Compared to a human? 

HOROWITZ: We have about five million cells.

DUBNER: So, does that mean they smell, “better than us,” like, 60 to one? Or does it not translate that easily? 

HOROWITZ: I think it’s probably wrong-headed for me to try to translate it in terms of, “They smell 100 times better than we do.” But we do know in some cases they can smell many, many more minute particles of an odorant than we can.  

*      *      *

Imagine if each detail of our visual world were matched by a corresponding smell. Each petal on a rose may be distinct, having been visited by insects leaving pollen footprints from faraway flowers. What is to us just a single stem actually holds a record of who held it, and when. A burst of chemicals marks where a leaf was torn. The flesh of the petals, plump with moisture compared to that of the leaf, holds a different odor besides. The fold of a leaf has a smell; so does a dew-drop on a thorn. And time is in those details: while we can see one of the petals drying and browning, the dog can smell this process of decay and aging. Imagine smelling every minute visual detail. That might be the experience of a rose to a dog. 

*      *      * 

DUBNER: Here’s a small question with a perhaps gigantic answer. But why dogs? Why are they the animal that have become the primary human companion? 

HOROWITZ: It might be that dogs partly chose us, right? So, somewhere out there, there’s a wolf podcast and they’re saying, “You know, why humans? What is it about them?” And they think, “Well, maybe humans were going somewhere.” 

DUBNER: God, I want to hear that podcast so bad.

HOROWITZ: Me too. But I think they partly chose us. There was a coincidence of mutual interest or mutual tolerance. And then, there were some things about wolves’ constitution that makes them good for pairing with us.  

*      *      *

So, you want to make a dog? There are just a few ingredients. You’ll need wolves, humans, a little interaction, mutual tolerance. Mix thoroughly and wait, oh, a few thousand years. 

Or, if you’re the Russian geneticist Dmitry Belyaev, you simply find a group of captive foxes and start selectively breeding them. In 1959, Belyaev began a project that has greatly informed our best guesses as to what we believe the earliest steps of domestication were. Instead of observing dogs and extrapolating backward, he examined another social canid species and propagated them forward.  

Though Vulpes vulpes, the silver fox, is distantly related to wolves and dogs, it had never before been domesticated. Despite their evolutionary relatedness, no canids are fully domesticated other than the dog: domestication doesn’t happen spontaneously. What Belyaev showed was that it can happen quickly. Beginning with 130 foxes, he selectively chose and bred those that were the most “tame,” as he described it

These “tame” foxes were allowed to mate. The tamest of those were mated, when they were old enough; and their young; and their young. After forty years, three-quarters of the population of foxes were of a class the researchers called “domesticated elite”: not just accepting contact with people, but drawn to it. He had created a domesticated fox. Incredibly, by selecting for one behavioral trait, the genome of the animal was changed in a half-century. And with that genetic change came a number of surprisingly familiar physical changes. They have floppy ears and tails that curl up and over their backs. Their heads are wider, and their snouts are shorter. They are improbably cute.  

*      *      *

DUBNER: When you write about why dogs fit so well with humans, you also make the point that physiologically, they’re not that unlike us. I mean, they are quadrupeds but they face forward. They don’t hop or slither or whatever. They walk with us, things like that. 

HOROWITZ: And there are a lot of things, absolutely, just about them that appeal to us. These are furry animals that have a face which was somewhat recognizable, right? And in fact, as we’ve bred dogs, we’ve made them even more recognizable with flatter faces, sort of more primate-like. And then they had this feature, at least somewhere along the history of domestication, that they would look us in the eyes in a way that actually wolves don’t usually do with each other unless they’re threatening each other. But that is super-meaningful for us. With dogs, it’s quite normal that a dog will sit there as you talk to them and look at you with their head cocked.

DUBNER: In part, because they actually do pay attention to us, don’t they? 

HOROWITZ: Yeah. I mean, not always with understanding, but certainly they’re paying attention to us. We are important to dogs. And I guess we like that feeling. 

DUBNER: You write that dogs are intuitive anthropologists, that they pay attention to us and continue to do so even after the novelty wears off — unlike many humans, you point out, who kind of stop paying attention to all but the biggest changes or the noisiest signals. So, does this mean that on some level, dogs are kind of more — I guess the phrase these days is “lifelong learners” — than many or most humans are?

HOROWITZ: I do consider them little anthropologists in the room watching our behavior and learning. We all get adapted to familiar environments and notice differences. That’s how the mechanism of attention works with us. And that’s what it’s for. But I think the difference is what counts as ordinary for them might be different than what counts as ordinary for us. So, I walk outside in the morning and if I’m looking at my New York City block it looks pretty much like it did yesterday. I’ll notice if there’s something fabulously different, right? If there are no cars at all, or if there’s a giant car, or if there’s a person standing right in front, I notice that. But that’s not what the scene is for the dog every time the dog comes outside, right? It’s a totally rewritten scene. All sorts of things have happened that have left sort of smell traces. 

DUBNER: For instance? 

HOROWITZ: Well, people have come by. Animals have come by. Garbage has been deposited. The air has changed. Maybe the weather has changed. And that’s brought scents from other places. Barometric pressure actually leads to an olfactory difference in the air.

DUBNER: Oh, really?

HOROWITZ: Well, have you ever gone outside before a thunderstorm and you’re like, “It smells like rain?” 

DUBNER: Once in a while. But I’ve been outside with my dog And she didn’t want to go far from the apartment at all. Like, we’d be going into the park off the leash. And she just didn’t, didn’t, didn’t, didn’t want to go. And we went home. And 15 minutes later it rained, and I had no idea it was going to rain. 

HOROWITZ: Right. Well, first of all, you just weren’t that perceptive, probably, Stephen. But to give some credit to your dog, I do think that what happens when barometric pressure changes — and that happens well before the storm is about to come — is that, if the pressure lowers, then smells that are in the ground can kind of come into the air.

DUBNER: And many dogs are not fond of the rain, yes?

HOROWITZ: I don’t know. I don’t think that’s necessarily true. There’s variance in personalities even variance within a breed. I would not say dogs don’t like the rain in general.

DUBNER: So, tell us why outfitting your dog in a slick yellow raincoat is maybe not the best idea.

HOROWITZ: Yeah, I think that’s the perfect example of our putting our experience and our feelings in the best-intentioned way onto the dog without considering whether it’s relevant for the dog. I don’t want to get wet, so I put on a raincoat or a rain hat. But our dogs are mostly pretty much fine with getting wet after they’ve gotten wet. And certainly, a rain jacket won’t prevent that feeling. And in fact, it introduces a new feeling, which is something pressing down on you. If your dog dresses themself in a raincoat before they go out into the rain, which they’ve smelled, then you should absolutely let your dog wear that raincoat. But otherwise, we’re forcing the dog. They freeze in place like they’re being scolded by an older wolf. And then, they endure it. 

DUBNER: Because you’re saying pressing down is a submission signal, yes? 

HOROWITZ: Well, being pressed down is a kind of dominant signal. And I would think that the raincoat probably reproduces a little bit of that feeling. I mean, they’ll put up with it but I don’t think it improves the situation for them the way it improves it for us to go out in the rain. 

DUBNER: I wanted to talk about a recent tweet of yours. You posted a photo from a stock agency, and it showed a little girl with a big smile hugging her dog. And the caption provided by the photo agency was, “Cute little girl hugging golden retriever with love, eyes closed, smiling.” And in the tweet, you suggested a more accurate caption would be, “Dog mightily managing self-control while being squeezed around the neck in an uncomfortable way.” So, can you talk a bit more about that? Does this mean, first of all, that dogs don’t like to be hugged, for instance — some dogs, all dogs, etc.? 

HOROWITZ: Yeah, I think the hug is a very human way of showing affection, right? And you’ll notice that very few dogs hug each other, that’s not a greeting that they do. Often dogs will be fine with it because they put up with a lot from us, right? That’s part of what dogs constitutionally are. They’re flexible and adaptable to different behavior toward them. But you never see a dog working to get inside of a hug of yours, right?

And that’s what a lot of these types of photos are about. Here’s a little person hugging a dog, and you just see the dog like, keeping it together with their head, looking over the side. And it’s scary because the next move of that photo could be the dog biting the face of the child and not meaning to be aggressive, but just trying to get out of this uncomfortable position. And then, suddenly, it’s an aggressive dog that has to be returned to the shelter or the breeder. I feel very concerned about photos like that because they give the idea that that’s the kind of embrace we should be in with our dogs. That’s the appropriate dog relationship. And that, to me, looks like a problem about to happen. 

DUBNER: If I put my dog in a tuxedo, let’s say, and she looks like she’s smiling in the tuxedo, is she smiling?

HOROWITZ: So, dogs don’t have muscular control of their lips. It’s not an actual smile. It’s like the dolphin smile where their face is just fixed in that position. It’s hard to not see a “smiling” dog as a happy dog. It’s our impulse. But that impulse, we should just be hesitant about, in making the attribution. In this case, it’s almost definitely not a sign of happiness. 

DUBNER: You do warn against anthropomorphizing our dogs too much — to see everything they do and think as having a human parallel. But that’s you as a scientist and a writer. What about you as a dog owner? Do you follow your advice? 

HOROWITZ: I do make attributions to my dogs that I think are anthropomorphisms, absolutely. I see Finnegan especially as being proud with like a large stick, for instance. But I also, at the same time, have the experience of saying, “You know, if you have a large stick, you have to hold your head up more because you’re trying to balance a stick.” So, I can’t really say anything about his experience. And that’s also okay with me.

I kind of like that there’s a mystery there, that it’s not just automatically what I assume it looks like. And I think anthropomorphisms actually have a use insofar as they get us interested in looking at dogs, and thinking about them as creatures who have experiences. And then, the question is, can we just suspend for a second the feeling that we know everything about them and just wonder or ask a question that can be tested instead of just assuming?  

*      *      *

I frequently hear dog owners verify their dogs’ love of them through the kisses delivered upon them when they return home. These “kisses” are licks: slobbery licks to the face; focused, exhaustive licking of the hand; solemn tongue-polishing of a limb. I confess that I treat Pump’s licks as a sign of affection. “Affection” and “love” are not just the recent constructs of a society that treats pets as little people, to be shod in shoes in bad weather, dressed up for Halloween, and indulged with spa days. Before there was any such thing as a doggy daycare, Charles Darwin (who I feel confident never dressed up his pup as a witch or goblin) wrote of receiving lick-kisses from his dogs. He was certain of their meaning: dogs have, he wrote, a “striking way of exhibiting their affection, namely, by licking the hands or faces of their masters.” Was Darwin right? The kisses feel affectionate to me, but are they gestures of affection to the dog? 

First, the bad news: researchers of wild canids report that puppies lick the face and muzzle of their mother when she returns from a hunt to her den — in order to get her to regurgitate for them. Licking around the mouth seems to be the cue that stimulates her to vomit up some nicely partially digested meat. How disappointed Pump must be that not a single time have I regurgitated half-eaten rabbit flesh for her.

Now the good news: as a result of this functional use of mouth licking — “kisses” to you and me — the behavior has become a ritualized greeting. In other words, it no longer serves only the function of asking for food; now it is used to say hello. Dogs and wolves muzzle-lick simply to welcome another dog back home, and to get an olfactory report of where the home-comer has been or what he has done. Since these “greeting licks” are often accompanied by wagging tails, mouths playfully open, and general excitement, it is not a stretch to say that the licks are a way to express happiness that you have returned. 

*      *      *

DUBNER: You do write about this one experiment, I’m pretty sure it was not one of yours — designed to tease out how dogs respond in an emergency, where the owners were instructed to fake a heart attack, or to pretend to have a bookshelf collapse on them. And you note that the dogs didn’t really do that much to signal for help. But you’ve also told us in the book that dogs sense a great deal about our inner lives by observing our behavior and also by sniffing out our excretions: our sweat and testosterone and maybe cortisol, whatever. So, I wondered, in an experiment like that, if the humans are faking it, wouldn’t the dogs kind of know they’re faking it or at least not be concerned that it’s real? 

HOROWITZ: Yeah, I think that’s a really good complaint about the experiment. And those experimenters, I think, realized it after the fact — that, “Oh, well, maybe the dog wasn’t responding because the person didn’t smell like they were stressed or were having a heart attack. So, why should the dog respond?” Another part of it is, even if they’re really sensitive to our physiological or behavioral changes, which I think they are, that doesn’t mean they always know the meaning of those changes.

So, a dog who comes to me if I’m crying isn’t necessarily doing the same type of thing that a person is if they come to me when I’m crying as a comfort. The dog may in fact be a comfort. But what the dog definitely knows is I’m doing something different. I’m making a lot of noise. I probably smell different. Like, let’s go see what’s going on here. And so, in the emergency example, even if they had succeeded at that test, it wouldn’t make sense to me to say, “Oh, that’s because the dogs understand that there’s an emergency situation.”  This type of research was generated from the kind of hero-dog narration where you read about in the news a dog that pulled three children from the water where they were drowning. And you think, “Oh my God, it’s amazing. I love that dog.” And then, people extrapolate and think all dogs are like that.

DUBNER: Or, as you write, people actually believe those stories 100 percent, when, in fact, there might be a lot of other factors going on in that hero-seeming situation. 

HOROWITZ: Exactly. Were they trained to do that? Did they just accidentally do it? Were the children actually okay? You can’t tell. And we also never get news reports of three children drowned because the dog standing on the bank failed to see that they had to be saved. We never get that news story, right? “Dog stands idly by on riverbank while people are drowned.”   

*      *      *

Today we’ve been getting inside of a dog with Alexandra Horowitz, a cognitive scientist at Barnard College whose first book is called Inside of a Dog. It turns out we make a lot of assumptions about dogs that are half-true at best. We took a microphone out to the streets of New York to ask some people with dogs what they know, or think they know, about how their dogs think and behave. Why, for instance, dogs urinate where they do. Here’s one response we got:

DOG MAN: They’re marking their territory, that’s like their whole — they’re kind of like putting little flags around the neighborhood to let people know it’s theirs. 

What’s Alexandra Horowitz think of that explanation?

HOROWITZ: It’s totally right to think of the marking behavior, just peeing a little bit on lots of different things, as leaving little flags, it’s sort of like little calling cards that say, “Me, me, me!” But it’s not territory, right? I mean, if dogs were really marking their territory, we’d expect that they’d go around the perimeter of your apartment or house and are marking every spot along the wall, because that’s their territory. You know, instead, it’s just leaving information about themselves in places that other dogs can sniff it. 

DUBNER: Alexandra, if I may I’d like to ask you about another behavior my dog does — because I’m sure you never get people asking you questions about their own dogs’ behaviors. I’m going to play a piece of tape. You’ll hear two sounds on this tape. One of the sound-making objects on the tape is my dog, and the other — well, it’s easily recognizable. Here we go: 


DUBNER: So, I know that my little Havanese is very far from her wolfy roots, but you do write in your book that among adult wolves, “A chorus of howling may help coordinate their travels and strengthens their attachment.” Is it possible that little 11-pound Fifi is trying to do something like that with the ambulance? What’s going on there? Any idea? 

HOROWITZ: Getting back to her ambulance roots, yes. Yes, I think that’s what she’s doing. I think that she’s kind of howling with the pack, essentially. And I think that’s totally charming, very delightful. 

DUBNER: So, my interpretation went a little bit deeper, and I’m sure it’s 100 percent wrong. But my interpretation was that, because she’s so brilliant, she learned that ambulances are transporting sick people and that she’s empathizing, calling ahead to the hospital. “Hey, take care of these people.” Do you think that’s true?

HOROWITZ: I mean, as I say, we readily make attributions to our dogs. 

DUBNER: How often are you asked to answer a question like this from someone who manages to get hold of you? 

HOROWITZ: All the time. It tells me a little bit more about you than it does about the dog, frankly. And that’s the case with most questions. 

DUBNER: So, dogs plainly can’t speak, at least not the way humans speak. But how much and how well do they communicate to humans, if you know what to listen and look for? 

HOROWITZ: A lot of things that we would ask of dogs, or dogs might ask of us, they’re showing us all the time with their behavior. If a dog needs to get my attention, he’ll do a number of things before he comes in here and barks at me. Finnegan, for instance, if he needs to go out or wants to play, he’ll come to my office door and just hang out there. So, it’s subtle, right? But I now recognize that that’s a thing.

DUBNER: So, is there one or a handful of things that you, as a dog lover and dog scientist, would most like to know about dogs, but because dogs can’t speak, you don’t yet know? 

HOROWITZ: Well, one thing that’s notoriously difficult to discover is the extent of their kind of autobiographical memories. Do they tell a story of their life, you know? Of course, they have perfectly good memories for places and people and things that have happened — but what shape it is. I certainly treat them as though there’s a “who” there. But what do they think of themselves? 

DUBNER: So, a lot of what you know about dogs’ interior lives comes from your own research lab at Barnard, the Dog Cognition Lab. Can you describe how a dog lab — or at least your dog lab — actually works?

HOROWITZ: Sure. Mine might be a little bit unusual in that there’s not a lab that’s constantly staffed. We work together through the course of a year on one or several studies en masse. So, we’re spending the year developing an idea that we want to test, developing a methodology with which to test it, recruiting subjects, and then having people come in with their dogs to run them through the experiments. 

DUBNER: So, the dogs are not living there. 

HOROWITZ: No dogs are kept at the lab. It sounds like we must have dogs in cages stacked on top of each other. All of our dogs are owned dogs. Most of them are New York City dogs. But sometimes, people drive in from hundreds of miles away to be part of the study. And they come with their people. The person is in the room with the dog. Our lab is physically just like an office that has nothing in it except for dog-related paraphernalia and some video cameras. And we run the dog through a little problem-solving test. And then, they go home with their person. 

DUBNER: And how representative do you think the dogs that you get in New York City are? 

HOROWITZ: It’s a great question because we certainly have two things that change our pool of dogs from sort of dogs generally. One is they’re mostly New York dogs. So, they’re dogs who live in an urban environment. Who are usually really well-socialized with other dogs and with people because they are around other dogs and people all the time. And then, on top of that, of course, the people who are interested in doing this are maybe not the average dog owner. Right.

DUBNER: Can I tell you, I once tried to sign up to bring my dog into your lab for an experiment and the paperwork required some information that I didn’t have handy. I think it was a vaccination record, like it needed a number or something. So, I had to wait until I went home that evening. But by the time I filled it out, I was told the experiment was already full. So, then, I was thinking, “Woah. At this lab, it’s selecting for the kind of dog owners who are not only able to immediately respond to the email solicitation on behalf of their dogs, but they also have their complete vaccination record at their fingertips.” And so, maybe these dogs, or at least the dog owners, are not super typical. 

HOROWITZ: I hope you’ve since corrected your ways, Stephen. Can you just tell me her vaccination record right now? 

DUBNER: I can’t. I feel like such a failure as a dog owner but I guess, again, this goes to the point of how representative the dogs, but more of the dog owners, are. So, your overall answer is what, representative enough? The best you can get?

HOROWITZ: You know, what we do is we’re just transparent, we’ll say, “Look, this is typical of dogs. We can probably generalize to other dogs like these dogs, right?” So, that’s all we can do. I mean, there are more stray dogs than owned dogs in the world. And I hesitate to say that I know everything about dogs because we’ve run one study with New York City dogs. 

DUBNER: So, in some ways, this is a parallel to a lot of different kinds of research that’s been done over the years. You know, a great deal of what we know about psychology and even economics in an experimental setting comes from basically college undergraduates at big research universities, and so on. Do you feel like there is any major component of dog cognition that you haven’t really gotten to yet because you’re mostly studying city dogs?

HOROWITZ: I think if I really want to have a sense of dogs’ perception of the world through smell, which is one of my great interests, I have to get the person out of there a little bit. I mean, it would be more interesting to study dogs who don’t have a person dragging them by the leash and don’t restrict where they can go. But just dogs being dogs and really track what they know by smell. Because people really restrict and control their dog’s environment. And so, you do get this kind of adulterated subject at some level. Again, that itself is interesting, right? These dogs don’t quite exist independently of the relationship with their people. So, we intentionally include the people in our studies.

DUBNER: But not getting the dog qua dog. 

HOROWITZ: That’s right. And I would be interested in that for sure. 

DUBNER: So, some dog breeds are thought to be naturally aggressive — pit bulls and Rottweilers and Dobermans. Are they? 

HOROWITZ: No, breeds are not naturally aggressive, period. And you know, the pit bull is the current bête noire. They certainly have a horrible reputation. Aggression is more complicated than that. Does aggressive mean that they bark at things that arouse them, that stimulate them? Does it mean that they attack somebody? What counts as aggressive behavior is really changed according to context. So, for instance, the reason that dogs like German shepherds were thought of as aggressive for so long is because they were used to attack, to bite people, to catch people. But is the breed therefore obviously always more aggressive? No, those dogs were trained to do that behavior, wouldn’t naturally have done that behavior, and there was nothing about the breed which is different. So, I think it’s how we’ve used dogs that’s led to their being tainted as aggressive. And then, they can’t ever get rid of that handle.

DUBNER: And then, it becomes a self-fulfilling prophecy — or it becomes a selection issue. People who want aggressive dogs get those breeds and teach them to be aggressive, you’re saying? 

HOROWITZ: At some level. And it’s also confirmation bias where we remember the cases of a pit bull who attacked a woman in the face or killed a child or another dog. And we don’t remember the case of the golden retriever, say, who bit a child in the face. Because we want it to be as simple as some breeds are aggressive and some breeds are not. But it’s not that simple. 

DUBNER: And you write, “Research found that of all breeds, dachshunds were the most aggressive to both their own owners and to strangers.” Little wiener dachshunds were the most aggressive? So, how was that measured and concluded? 

HOROWITZ: So, this was based on a very large survey of dogs that’s been going on for maybe a decade and a half, the C-BARQ. 

DUBNER: That stands for something? 

HOROWITZ: Yeah. I’m sure it does. And it begins with “canine” and ends with “questionnaire.” But I don’t remember the middle. And it’s a battery of questions that owners, that dog people answer about their charges. And then, the researchers look at, you know, are there differences that we can see by breeds, by sex, by age, etc.? And they found that in terms of owner-reported aggressive acts — things like barking, biting, snarling, growling — dachshunds showed far more than the other breeds. And then if you go around looking at dachshunds and thinking, are you an aggressive little dog? They’re probably barking at you. Now, are they, in fact, aggressive in this other way we mean? Like, can they hurt me? You can pick up the dachshund, stash them in your tote bag, and then, they’re not aggressive anymore. That’s harder with a German shepherd.

DUBNER: So, should we conclude that it’s good that dachshunds aren’t much bigger, in which case they could do damage? Or should we conclude that dachshunds are aggressive because they’re small and they just need to kind of make themselves known? 

HOROWITZ: It would be terrifying to imagine a really large dachshund, wouldn’t it?  

DUBNER: It’d be like one of those stretch S.U.V. limos that you see on prom night. 

HOROWITZ: I think that would be a failure as a breed. But yeah, it’s reasonable to hypothesize but not conclude that they’re doing that behavior because they’re small dogs, right? I think it’s maybe a compensatory element. You know, in all animals, including the human animal, there’s something — position that you gain by being larger or louder. And so, if you’re small, maybe you could be louder. 

DUBNER: But you do write that small dogs don’t think of themselves as lesser than big dogs, correct? 

HOROWITZ: Right. I don’t think they think of themselves as weak or of a different species than a big dog, right? In all my play studies, small dogs are perfectly good at playing with giant dogs and vice versa. The large dogs moderate their behaviors to be more suited to the small dog. They don’t just eat them. 

DUBNER: When you think about the symbiosis of the dog-human relationship, who do you think gets the better end of the deal? 

HOROWITZ: I don’t know that I totally view it that way, right? But I do think that we’re getting a better deal in some ways. Our need to take kind of total control in many cases over the dog’s experiences — when they go out, when they eat, when they can poop, they can deal with it. But I feel like there are many fewer controls that they have on our experience. So, I think we’ve kind of got the better deal. 

*      *      *

By standard intelligence tests, the dogs have failed at the puzzle. I believe, by contrast, that they have succeeded magnificently. They have applied a novel tool to the task. We are that tool. Dogs have learned this — and they see us as fine general-purpose tools, too: useful for protection, acquiring food, providing companionship. We solve the puzzles of closed doors and empty water dishes. In the folk psychology of dogs, we humans are brilliant enough to extract hopelessly tangled leashes from around trees; we can magically transport them to higher or lower heights as needed; we can conjure up an endless bounty of foodstuffs and things to chew. How savvy we are in dogs’ eyes! It’s a clever strategy to turn to us after all. The question of the cognitive abilities of dogs is thereby transformed: dogs are terrific at using humans to solve problems, but not as good at solving problems when we’re not around. 

*      *      *

DUBNER: So, dog ownership is way up during the pandemic. How do you feel about this new army of dog owners? And how do you feel about that word, “owner,” applied to the humans that a dog lives with? 

HOROWITZ: I worry about ownership generally, right? I wonder if that’s the relationship we should have with our dogs generally. But I think it’s wonderful that people turn to dogs as the kind of proxy for human company that they weren’t getting in this time of being isolated. And it was terrific for the shelters. You know, it cleaned out a lot of shelters. There is, of course, an asterisk on this. We worry about what’s going to happen to these dogs when the people are going back to work. Are they all people who adopted dogs with a kind of full understanding of a life with a dog, even if you’re not working from home? 

DUBNER: So, you think the shelters might fill up again in a year or two? 

HOROWITZ: Yeah, I think it’s absolutely possible. And I know, in some cases, shelters already have had a lot of returns of dogs.

DUBNER: So, if someone is on the fence about getting a dog, let’s say they’ve never had a dog, how would you advise they consider making the final decision?  

HOROWITZ: Well, it’s easy to get a dog. But what you’re actually signing up for is if you’re lucky, a 20-year relationship with a living being whose needs you can’t anticipate yet. None of us is prepared for how it’s going to change us and what we actually need to do to treat that dog well. So, spend a little time girding yourself for the fact that if you’re doing it right, your life is going to change. And that’s not what you think you’re signing up for often when you think, “I’m just gonna get a dog. This is such a cute puppy.”

DUBNER: If there was just one or two things that you would want everyone to think about as they’re getting their dog for the first time to try to really understand a dog in a way that typically isn’t talked about so much, where would you start? 

HOROWITZ: Well, I would say not having preconceptions about who that individual dog is or can be. I think that’s really important and very tough, right? When you decide to buy or adopt a dog, you already have an idea of where they’re going to fit in your family. Well, maybe they don’t. Like, they’re a different person. And being open to that would save a lot of aggravation. And I think the next thing is — and I don’t study this per se, but I think it’s important — thinking about what the role of training is in your dog’s life. 

I’m not a big advocate of just training for training’s sake. You know, who cares if your dog can sit on command? I don’t care. The reason you teach something like that is on the way to forming a good relationship with your dog. Teaching your dog “sit” is not about your dog sitting. That’s not a trick that you should pull out and feel satisfaction for. Instead, it’s sort of the beginning of your dog understanding that there are some times you’re going to need them to be still and that they can actually control their own body. So, that’s like the beginning of an understanding. It doesn’t end with learning to sit or lie down or rollover. That’s not the endpoint. The endpoint is a relationship. 

*      *      *

We and our dogs come closer to being a benign gang than a pack: a gang of two (or three or four or more). We are a family … Trainers who espouse the pack metaphor extract the “hierarchy” component and ignore the social context from which it emerges. A wolfcentric trainer may call the humans the pack leaders responsible for discipline and forcing submission by others. These trainers teach by punishing the dog after discovery of, say, the inevitable peed-upon rug. The punishment can be a yell, forcing the dog down, a sharp word, or jerk of the collar. Bringing the dog to the scene of the crime to enact the punishment is common — and is an especially misguided tactic. 

Punishing the dog for his misbehavior — the deed having been done maybe hours before — with dominance tactics is a quick way to make your relationship about bullying. The result will be a dog who becomes extra-sensitive and possibly fearful, but not one who understands what you mean to impart. Instead, let the dog use his observation skills. Undesired behavior gets no attention, no food: nothing that the dog wants from you. Good behavior gets it all. That’s an integral part of how a young child learns how to be a person. And that’s how the dog-human gang coheres into a family. 

*      *      *

DUBNER: I want to know how much a given dog’s temperament reflects the personality of the humans that they live with. So, I play with my dog a lot, we do a lot of chasing and fetching and hiding and seeking, etc. But I want to know, is she playful because I play with her a lot? Would she be playful if her human didn’t play with her a lot? How much symbiosis is there in the personality/dogonality dimension? 

HOROWITZ: First of all, I just love imagining you like, hiding with your dog. That sounds like a lot of fun. And I think we find the things that enable us to work best together, right? So, if you have a certain kind of play you like to do together, then you will slowly find yourself doing that type of play all the time. Did it happen in you first? Is it constitutional of you or did it happen in her first? Is it constitutional for her? I don’t think that’s the right question, really. You can really just talk about, “you’re playful together in this dyad. “Would she be as playful with another person or dog? I don’t know. Because it isn’t necessarily just a part of her. It’s part of her-you, right? And I think that’s what’s interesting about dogs, unlike other animals, is that there’s a way that we’re really appended together. And you almost can’t talk about the owned dog in separation from the person.

DUBNER: I wondered about that, because you write how we have an effect on our dogs, but our dogs have an effect on us. And for me, the effect is 99 percent salutary. I grew up with dogs. I love dogs. And I love the dog that I have now. And, until I read your book — or reread your book for this conversation — I hadn’t really thought about what effect she’s had on me overall, especially in my work life. She’s 7 now. And since we got her, she’s come with me to my office or wherever I’m writing pretty much every day and spends all day with me. And I never really stopped to think how might she have “influenced my work” — I mean, changed the way that I think and feel and move through a day.

Some of the things are concrete, temporal things. Like, every day there’s a walk or two which I wouldn’t take otherwise. And there are encounters on that walk. But there’s also just a feeling of companionship that I’d love to know, if I could, how that’s affected everything I’ve written and talked about for the past bunch of years. I’m curious if there’s any research on how a dog in daily life affects your trajectory, temperament, and so on. 

HOROWITZ: The only kind of research out there is really ones that look at, for instance, levels of oxytocin production. When you lock eyes with a dog or when you touch a dog, there are studies looking at the fact that your oxytocin levels go up.

DUBNER: And that’s a good thing, we should say. 

HOROWITZ: It’s this peptide hormone, yeah, that’s involved in feelings of affection and attachment. And so, that seems absolutely health-giving on a daily basis and would kind of put you in a better place, just as being in a great relationship with a person might put you in that place.

DUBNER: So, what are you writing now? 

HOROWITZ: I’m writing a book on early dog development, which is tentatively titled Year of the Puppy.

DUBNER: And do you think you will be writing about dogs when you’re 80? 

HOROWITZ: I would be so lucky to be writing about dogs when I’m 80. 

DUBNER: So, here’s a short passage from your book that really intrigued me. “Dogs are ingenuous. Their bodies do not deceive, even if they sometimes cajole or trick us. Instead, the dog’s body seems to map straight to his internal state, their joy when you return home, or their concern, plotted by the lift of an eyebrow.” So, considering this ingenuousness and the fact that dogs don’t speak and if you can’t speak, you can’t tell a lie — what does this say about dogs’ honesty or decency? Does it suggest that we should all on that level strive to be a little bit more doglike?

HOROWITZ: I do think they are fundamentally decent. It’s the perfect word to describe dogs, and not only that, but unrelentingly cheerful in the face of non-cheery situations, right? And I do try to model myself on that behavior. Or at least nudge myself in that direction and tone down the skeptical, grumpy part of myself. I mean, how fantastic that they don’t hold a grudge against you when you’ve left them? They’re so delighted that you’re back. If I’ve ever met a person who did that I’d like, capture that person and stay with them forever, because it’s a rare trait indeed. 

That was Alexandra Horowitz, proprietor of the Dog Cognition Lab at Barnard, and author of Inside of a Dog: What Dogs See, Smell, and Know. Her other books include Being a Dog; Our Dogs, Ourselves, and On Looking. They can all be acquired wherever you acquire your books these days. Please let us know what you think of this Freakonomics Radio Book Club format — we’re at We’ve got a few more episodes already in production, including a book called Eat Like A Fish: My Adventures Farming the Ocean to Fight Climate Change, by Bren Smith. So, if you’re the kind of person who likes to read ahead — go for it.

*      *      *

Freakonomics Radio is produced by Stitcher and Dubner Productions. This episode was produced by Brent Katz. Our staff also includes Alison CraiglowGreg RippinMary DiduchCorinne Wallace, Zack Lapinski, Daphne Chen, and Matt Hickey. Our intern is Emma Tyrrell. Our theme song is “Mr. Fortune,” by the Hitchhikers; the rest of the music was composed by Luis Guerra with additional music this week by Michael Reola and Stephen Ulrich. You can subscribe to Freakonomics Radio on Apple PodcastsStitcher, or wherever you get your podcasts.

Read full Transcript



Episode Video