[Audio: Dog whining]
Lynne Malcolm: That’s Daisy talking, she’s a petite and pretty Dalmatian, with one blue eye and one brown eye. She’s trying very hard to tell her owner Nio something.
[Audio: Dog whining]
But what is Daisy thinking?
Hi, you’re withAll in the Mindon RN, I’m Lynne Malcolm. Today, what’s in the mind of your pet dog?
This is what Gregory Berns wanted to find out about his dog, and he went to unusual measures to do it. He’s a Professor of Neuroscience at Emory University and I caught up with him on Skype.
Gregory Berns: I have to warn you, there are dogs here.
Lynne Malcolm: There are sound effects readymade. How many dogs have you got with you?
Gregory Berns: There are as many as four somewhere around the house.
Lynne Malcolm: Gregory Berns, who’s running what he calls the Dog Project.
Gregory Berns: The main thing that I wanted to find initially when I started the project was what do dogs think of us humans? So the key question for me has always been do dogs hang around humans because we feed them or do they actually have feelings and something analogous to love, or do they like us just for the bond itself?
Lynne Malcolm: So he decided to investigate by scanning the brain of his pet dog, Callie, in an MRI machine.
Gregory Berns: For most of my career I’d spent roughly 20 years actually studying human brains and how humans respond to various things like incentives and risks and rewards, and the Dog Project just started as a crazy idea of first asking the question could I train my dog to go in an MRI scanner so I could see what she was thinking? And initially I didn’t really have any expectations that it would even work. We weren’t sure that we could train a dog to do this, that we could do it in a way that they would like. And in fact it turned out to be not too difficult. It only took a few months to train my dog, and it kind of exploded from there. We started with two dogs and actually just this last week we successfully scanned the 100th dog.
Lynne Malcolm: So to have an MRI is quite a traumatic thing in a lot of ways, it’s very noisy and you have to go into a closed space and you have to be very still. Just give me a sense of how you trained these dogs.
Gregory Berns: So most people do not enjoy getting an MRI scan, and the things that bother people are not necessarily the things that bother dogs, but some things are in common. So people get claustrophobic, which turns out not to be a big deal for dogs. But the thing that I think bothers everyone is the fact that they are quite loud. So I approached a local dog trainer that I knew to help with this project to see if he was interested, and we just started breaking down the elements of an MRI scanner. I actually just started building parts of them in my basement to simulate it. And the whole idea was to essentially make it a fun game for the dogs. So my dog was the first dog and so I would just play with her while I was playing recordings of the sound, and that’s how she got used to it.
Lynne Malcolm: Gregory Berns emphasises that the dogs enjoyed the game he created to keep them in the scanner and they were free to walk out at any time, and only a few did.
Before we hear more about his findings, let’s meet some dog owners.
Nio: Hi, my name is Nio and this is my dog Daisy. Daisy, come on. What are you doing? No, down, off the couch. She is a liver spotted Dalmatian. She’s very small, she was the runt of the litter and I think she kind knows that she is always been the little left-out dog, so she is very attention-seeking all the time and always tries to get your attention.
Lynne Malcolm: Yes, she is putting her paw up on the couch right now, she just wants to be with us all the time, doesn’t she.
Lynne Malcolm: Do you often wonder about what she’s thinking?
Nio: Yeah, definitely. I definitely think she knows what humans are feeling because if I am having a bad day and I’m really tired she tends to not get up and kind of wait before she gets crazy and wants to go out for a walk. But if she knows I’m up and I’m kind of up and about and have a lot of energy, she can feel it and she instantly wants to go out. Yes, I always wonder if she can form those thoughts in her mind thinking, no, she is really tired, I might just wait, or if it’s just something that happens automatically for her.
Lynne Malcolm: Do you think she loves you?
Nio: Yes, I do, but I’m not sure if she kind of loves me particularly just because I am me, I think it’s just what people give her, she gives back sort of thing, because she loves strangers too. If someone has food she’ll run up to them and be just as cuddly as she is with me. I definitely think she feels love but I don’t know if she can distinguish love between different people.
Lynne Malcolm: Nio, with her dog Daisy.
Leni: Lupie, come! Sit. Good girl. Roll over. She just wants a treat. Roll over, roll over. Good enough, okay.
Hi, I’m Leni.
Ben: And I’m Ben.
Leni: This is Lupa. We got her from a pound about two years ago, and we think that she might be about three or four years old, and she is bull-Arab-cross-something.
Lynne Malcolm: She looks pretty happy at the moment. She’s got her chin resting on your knee and you have a little treat in your hand, she’s pretty keen.
Leni: Yeah, she is making full eye contact with the treat. She’s pretty food oriented.
Lynne Malcolm: So she was a rescue dog. Ben, tell me a bit about her character, what’s she like?
Ben: So she is generally pretty nervous around people and anything new, loud noises, strange objects, plastic bags, brooms, anything out of the ordinary she is quite cautious around, to the point where she gets so nervous that she’ll break down into shuddering and having convulsions.
Lynne Malcolm: Have you noticed a change though, an improvement in how she is?
Leni: She has definitely come to be very comfortable, she’s incredibly comfortable with us, very dependent on us, and she follows us everywhere, and also with people that we see frequently.
Lynne Malcolm: And do you think she loves you?
Leni: I think so, yeah, I think about that a lot, and obviously we provide her with food and water and shelter and warmth, and I think for a dog that probably is love. That’s how I see it I guess, is that we are nurturing her and she is appreciative of that nurture. And she definitely shows concern when…well, she shows concern when we go away, when one of us is leaving.
Ben: She’s definitely very tuned in to our emotions. Whether she loves us I’m not sure. It makes it easier to think that she does, to put up with all of the getting up at 6 o’clock in the morning and walking two hours a day and all of the things that come with having a dog.
Lynne Malcolm: Ben, Leni and their dog, Lupa.
Gregory Berns from Emory University used brain scanning to try and find out what dogs think and feel. But whilst you can look at brain structure and blood flow with an MRI machine, what can scans tell us about the actual experience of the dog?
Gregory Berns: That’s a great question and it’s a very difficult one to answer. So first you have to believe that the brain is the sole organ in our bodies responsible for our mental experiences. If you accept that, then it’s a mystery in the sense that we have to figure out what the different parts of the brain do and how that relates to these subjective experiences that we have. So we rely very heavily on knowledge coming from the human literature because only humans can report out what we experience.
So our very early experiments with the dogs were all based on human experiments, so our very first experiments were focused on what we call the reward system of the brain, and then the idea was to do things and show things to dogs that we know that they like and are similar in terms of what humans like. So we started with food. And the idea was to show hand signals to the dogs that signalled that they would get a piece of food. And what we got from that is we demonstrated first that, yes, the dogs will hold still in the scanner, that they will do this and enjoy it, and we can identify activity in their reward systems to these things that we know that they like.
Then we can start to build on that, we can make our experiments more complicated, we can show more complicated things to the dogs and always reference the similar experiments that are done in humans to interpret the idea that when we see similar activity in a dog’s brain as we do in a human’s brain we can reference the human experience to kind of get an idea of what the dog may be experiencing. So if we see the same patterns in the dog brain as the human, then we conclude that they may be experiencing something similar.
Lynne Malcolm: Gregory Berns and his colleagues began their research by looking at what dogs experience as emotions, and even love.
Gregory Berns: The very first things that we studied were essentially just the reward system, and we did that not because there’s anything surprising about the fact that dogs like food, but it gave us a map to start with. So after that we did a follow-up experiment where we showed cues that either signalled that the dog would get a piece of food, like they did in the original experiment, or a different cue that said that their owner would pop interview and just praise them, say something like, ‘Good girl,’ or ‘Good boy.’ And then the idea was to use the specific part of the brain as a gauge to see if they actually liked food more or whether they liked the prospect of being praised.
And it turned out that for most of the dogs, the vast majority of them in fact had equal responses to the prospect of food as well as praise, indicating that yes, they do actually like the social bond as much as we do probably. So if we need to put a label on that, we might be tempted to call that love. I’m perfectly okay with that label because it’s kind of an all-encompassing term that describes many different types of love, but we do know that the dogs have this response just to the social bond itself.
Lynne Malcolm: You’re withAll in the Mindon RN, I’m Lynne Malcolm. Today we’re investigating the mind of your pet dog with Professor Gregory Berns. As a neuroscientist from Emory University, he decided to learn more about being a dog by scanning their brains in an MRI machine. He’s written about his work in his bookWhat It’s like to Be a Dog.
When you come home and find your pet dog has chewed up your shoe, or has stolen a piece of cake off the bench and you reprimand them, do you think they seem guilty?
Gregory Berns: We have not studied guilt, and guilt is a very complex emotion.
Lynne Malcolm: Gregory Berns.
Gregory Berns: There’s actually not a whole lot of information even about how human brains represent guilt, but I will point out that it’s cognitively very complex, because to be guilty of something you have to remember that you did something, you have to remember that it violated some rule or some social norm, and you have to realise that somebody else is likely not happy with what you did. So it kind of takes all those pieces to get that feeling of guilt. So if you lack the ability to do any one of them then it’s not the case that you can feel guilty. So then we have to kind of back into…and ask the question do dogs have all those abilities, so we don’t know.
So just taking the memory question alone is that a dog would have to remember that they did something. There’s a fair bit of data that says that dogs can remember things for periods of 10 to 30 seconds, and then after that it becomes very unclear what dogs remember, at least in terms of discrete events. So if it’s the case that you come home and you see something chewed up on the floor, it’s not at all clear that the dog will remember that they did that. They might just see it and will just be lying there and wonder the same as you; who did that? So you have to take that into account.
It is clear though that dogs are very sensitive to how we speak and the emotions that we portray, and they certainly get when we are angry. So if a person reacts negatively to seeing something chewed up on the floor, well yeah, of course the dog is going to look like they are scared and they’re anxious because they know you’re unhappy and that something bad is likely to happen. It’s not clear if it’s guilt or they are just reacting to the person’s emotional state.
Lynne Malcolm: Gregory Berns.
Leni and Ben got their dog Lupa from the pound. She was a rescue dog. They think a lot about what emotions Lupa has.
Ben: Yes, unfortunately fear is the thing that I see most in her that is the most striking emotion. She’s so nervous, it’s painful to watch that fear that you just…it’s not fear in a person that you can rationalise out of them, you can’t tell them it’s going to be okay, you can’t talk her through it. And just seeing her like that is quite distressing.
Lynne Malcolm: Yes, I guess it really makes you wonder about her past, and there’s obviously some sort of trauma in her past.
Leni: And I guess on the flip side of that, seeing the fear, is also you can really tell then when she is really relaxed and when she is really happy because she will just settle down, she’ll curl up and go to sleep at your feet or she will be playful and she’ll roll around and she’ll submit herself, which takes a lot for her to do. Like right now she has gone and found a little bed for herself somewhere on the rug over there.
Lynne Malcolm: Yes, she is taking the opportunity to go and curl up.
Leni: Yes, she is very relaxed. So yeah, seeing her in her element and seeing her really comfortable is a really rewarding thing for a dog that has been in a pound.
Lynne Malcolm: Leni and Ben.
You’ve mentioned how important it is to recognise the individual differences in dogs and dogs’ characters I guess, and just as an example, one of the dogs I know is a rescue dog, and she is full of anxiety and fear and tentativeness, and another dog is almost the opposite, originally the runt of the litter, naughty, playful, and appears to have much less impulse control. What is this likely to say about their upbringing and their previous experience?
Gregory Berns: That’s a great question, and it’s actually quite difficult to disentangle these things, the same with people. For 50 years we’ve been struggling to understand nature versus nurture in humans. We did a study, probably the largest study now with dogs and MRI. We actually studied this question in a population of dogs training to become service dogs, assistance dogs. So what we did was we teamed up with a service dog organisation and we helped them train 50 of their dogs for the MRI as part of their regular training. And so the idea was to use the MRI to see if we could predict which of those dogs would actually become service dogs at the end of the process.
And it took two years to do it, but in the end we were able to hone in on a couple of key regions in the dogs’ brain, including the reward system, as well as a structure called the amygdala which is closely related to arousal, what we call physiological arousal or excitement. What we found was that the dogs who had more activity in this structure called the amygdala associated with excitement and arousal turned out to be the worst service dogs. So this was kind of the first hint, at least for me, that we were able to detect these low-level aspects of emotional responses in the dogs’ brain before it even becomes a behaviour, before they even show outward excitement or fear.
But back to your question of whether it’s the genetics or the upbringing, what was fascinating to me about that study is the dogs were all very genetically closely related because they all came from the same line of dogs, they were all golden retriever Labrador mixes, and they were all raised in the same environment, and yet there was still this tremendous variability there. So to me that kind of points more towards the genetics with the dogs, but we know it’s not 100%.
Lynne Malcolm: Gregory Berns.
Nio: Daisy, come on. No, down, off the couch.
Lynne Malcolm: She just keeps getting up on the couch.
Nio: Yeah. She is very curious, so she wonders what the microphone is.
Lynne Malcolm: There’s one situation that I’ve wondered about, what dogs are thinking when, for example, you know they desperately want to get up onto the sofa and they know it’s not allowed, what are they thinking, what’s stopping them from getting on the sofa? Which brings us to impulse control in dogs. Tell me about how you tested for that.
Gregory Berns: The experiment was based on, again, human experiments, and these were things that had been done for decades, originally in very young children. And this task, it’s called the ‘go, no-go’ task. In the human version typically what you do is you have a person, say, in front of a computer and you will maybe shown them letters on a screen, and the person’s task is to, say, hit a button every time they see an X. So you set it up in such a way that you show lots of Xs, so they are hitting the button, they hit it, X, X, X, X, and then something else comes up, say like the letter Y, and they have to put on the brakes and not hit the key when they see this new letter, that’s the no-go or ‘don’t go’ condition.
And so we call this a task of impulse control because we set it up in such a way that the person is used to hitting the target or hitting the button and then all of a sudden they have to not do that. So if you set these tasks up properly it really takes a fair bit of mental effort to put on those brakes.
So we did the same version with the dogs, except it was actually quite a bit more complicated to teach them to do something like this. We used a dog whistle. And so what we did was we first taught the dogs that when we blew the dog whistle, they had to nose-poke a little plastic target that we placed in front of their nose. And they were taught to do this in the MRI scanner actually. So that part of the task is actually fairly easy to teach the dogs to do. After they got good at doing that, then we introduced a hand signal which was the person raising their arms in kind of a cross shape, and that was meant to tell the dogs ‘don’t poke the target, even when you hear the whistle’. So that’s the putting on the brakes bit. And that part of the task took probably about six months to train the dogs. It’s extremely difficult to teach a dog not to do something. It’s much easier to teach them to do something.
But anyway, when the dogs finally learn that then we took them back to the scanner and had them do this whole task in the scanner, and what we were looking for was what part of the brain becomes active when they are putting on those brakes? And what we found was really quite amazing, it was a part of their frontal lobe, and it turned out it’s the corresponding part of their frontal lobe that is the same in humans, doing that particular task. What’s even more amazing is the dogs who had more activity in that area were the dogs who did best on the task. So it’s like the dogs who kind of had the neural real estate, if you will, to deploy on it were the ones who were actually doing well.
Lynne Malcolm: Gregory Berns.
Nio thinks Daisy is not naturally high on the impulse control scale, and she’s got plenty of funny stories confirming this. Here’s one:
Nio: So Daisy is quite a special looking dog, she has one blue eye and one brown eye, she always gets a lot of attention. So some people saw her and wondered if she wanted to be in an ad for Peter Alexander, and I said, oh, that would be really fun but sometimes she is very naughty, even though she behaves one minute, she sees something and she gets very excited, like Dalmatians do. So everything went really well, we had the photo shoot and she was laying down and we had a big break. It was in Centennial Park and I let her loose for a while and just let her get rid of some energy for the next part of the commercial.
And then in the end she saw a duck across the pond and she just ran for it, she went straight into the pond and tried to get this bird and I thought, oh my God, she is completely wet. Luckily no one saw her jumping in, so I kind of snuck out of the photo shoot and like, Daisy, come back! But then she got stuck in the lily pond and she got tangled up. My God, I have to now jump in and swim after her and get her. But in the end she managed to de-tangle herself and come back but she was completely brown with mud, and I just ended up having to take one of my sweaters and wipe her off before I took her back to the set because they would be mortified having this brown dog in a children’s pyjama commercial.
Lynne Malcolm: Nio.
So if there was one thing that you could find out about what Lupa thinks, what would it be? Is there a situation where you just think what is going on in her head right now?
Leni: I guess sometimes when she is in the car, when we take her for long car rides, she is so keen to get in the car. On the one hand it seems like that’s her safe space and we will open the door and she’ll just jump straight in. And then we will be driving and she’ll be shaking, like just completely shaking. And if she is in a scary situation outside and we open the car door she can’t wait to get in, but yet she is really anxious. So in those times when she seems to be displaying conflicting responses to a certain environment, I would love to know what she is thinking then.
Ben: Yes, seeing that inner conflict in her for me is really confusing. She will see people that she knows and is really comfortable around and she’ll try and go up to them, tail is wagging, she is visibly happy, bouncing around, but she will get within a metre of them and then back away and the tail will tuck for a few seconds until she is 100% sure that it’s the person that she thinks it is, and only then will she go up and be affectionate to them. So for me it’s wondering why that happens, if that’s a traumatic thing or if it’s just a nervous dog thing. But I’d love to know what caused that.
Lynne Malcolm: Leni and Ben.
Professor Gregory Berns says that his Dog Project research has really influenced the way he thinks about the ethics of the way we treat animals.
Gregory Berns: I don’t see how you can’t be affected by it. So the more I study dogs and how they react and how their brains react, the more similarities I see and parallels I see to humans. I see the same circuits being involved, the same emotional circuits, the same impulse control circuits. And you realise they have a lot of the same hardware that we have in our heads, and we have language and that’s not to be discounted, but language isn’t everything. And just because other animals can’t speak to us, that doesn’t mean that they not feeling things very similar to how we feel. And certainly when I see that, it has changed my view of animals. I don’t eat meat nearly as much as I used to and really try not to. And it changes how I view them as tools in medical research. The way that I structured this project is that it treats the dogs as volunteers, not as essentially slaves to the experimenter. It’s all voluntary. If the dog doesn’t want to be in the scanner, they walk out.
Lynne Malcolm: Gregory Berns, neuroscientist and distinguished Professor of Neuroeconomics at Emory University. He and his team have scanned 100 dogs now, so they have plenty of data for future research. They’re interested in finding out more about what makes one dog different from another, what they are feeling minute to minute, and what dogs actually get from human speech.
His bookWhat It’s like to Be a Dog: And Other Adventures in Neuroscienceis published by One World.
Thanks to Nio, Leni and Ben, and Daisy and Lupa.
Thanks also to the team, producer Diane Dean and sound engineer Andrei Shabunov.
And you might like to check out the extended version of this program on our podcast.
I’m Lynne Malcolm. Thanks for your company, bye till next time.