Science Friday - DNA Privacy, Dog Cognition. May 4, 2018, Part 2

Episode Date: May 4, 2018

Genetic testing sites are nothing new. They’ve grown enough in popularity over the past decade that the idea of spitting into a tube and sending it in the mail to a website to find out more about your family tree—or even your risk of certain inherited diseases—doesn’t seem all that strange to most people. But the case of the Golden State Killer has brought to light many questions about the direct-to-consumer genetic testing market that still need answering. Dr. Amy McGuire, professor of biomedical ethics at Baylor College of Medicine discusses the risks we take when we share genetic information online. Plus, Natalie Ram, assistant professor of law at the University of Baltimore School of Law discusses how this new era of genetic research is butting up against the criminal justice system. Sit. Come. Stay. Your dog knows how to do it all, and she even seems to understand what you’re saying. But every dog owner has probably wondered what exactly is going inside the mind of their prized pooch. Does Spot really understand what you’re saying, or is he just trained by the treat bag? Does Fluffy have a concept of time? And how do our furry companions make sense of the world? Neuroscientist Gregory Berns has trained dogs to sit inside fMRI scans to see what happens inside their brains. Subscribe to this podcast. Plus, to stay updated on all things science, sign up for Science Friday's newsletters.

This is Science Friday. I'm Iroflato. A bit later in the hour, we're talking all about dogs. What can neuroscience tell us about how they think and feel about the world around them? You want to talk dogs? Our number is 844-724-8-255. But first, genetic testing sites, they're really nothing new. They've been growing in popularity over the past decade. Enough so that the idea of spitting into a tube and sending it in the mail to a website to find out more about your family tree or even, you know, Your risk of certain inherited diseases, well, that doesn't seem all that strange to most people, does it? But in the case of the Golden State Killer, that has brought to light many questions that still need answering about direct-to-consumer genetic testing and family search databases. For example, who controls your data once you've shared it with a third-party site? Should law enforcement be able to use a consumer genetic database to aid in a criminal case? And when that data links you to a family tree, how do you protect the privacy of those that haven't willingly shared it like you have? That's what we're going to be talking about, our number 844-8255. You can also tweet us at SciFri.

Let me introduce my guest. Dr. Amy McGuire, Professor of Biomedical Ethics, Director of the Center for Medical Ethics and Health Policy at Baylor College of Medicine. Welcome to Science Friday. Thank you. Hi, Ira. Hi, Natalie Uram, assistant professor of law at the University of Baltimore School of Law. Welcome to Science Friday. Hi, thanks for having me. You're welcome, Natalie.

Let me ask you first. Can you take us briefly through the details of this case, the Golden State Killer case? How did the police come to use consumer genetic database to catch you? So the Golden State Killer committed at least a dozen. murders and maybe more than 50 rapes during the 1970s, and so it was quite a prolific criminal. And police had collected a lot of genetic material left over at those crime scenes, and over time had basically sequenced that genetic information, compared it to the genetic profiles of individuals in the California and the national known offender registry, a known

offender database of genetic profiles, and they did not get a match. And so they developed more detailed information about the Golden State Killer's genetic profile and started comparing it with publicly searchable genealogical databases. There was an unsuccessful effort where it looked like there might be a partial match, and and investigators obtained a judicial subpoena to get one individual's DNA and then a warrant for a direct DNA sample from someone who was not a match. And then investigators uploaded this profile of the Golden State Killer's genetic information to a website called jedmatch.com, which is designed to aid in genealogical research. And there they obtained, you know, partial matches that indicated not that the people who'd upload, their genetic information to Jedmatch had committed these crimes, but that a genetic relative

of one of those individuals might have committed the crimes, worked backwards from there, developing family trees, zeroed in on Joseph James DeAngelo, and then followed him around, collecting items he'd thrown away to obtain a discarded bit of DNA to make their final match. And then they arrested him. Let's just clear up a little bit when we talk about genetic websites. there are some sites that actually test your data, right, when you spit in the tube, when you send it out. And then there's other sites where you just upload your data to compare and connect with other family trees, for example. Right, Natalie?

That's correct. So a website like 23 and Me are an ancestry DNA service can both sequence your genetic information. That's who you send your, say, your cheek swab or your saliva sample to. They're going to return the data to you. They also allow you to download a much more complete set of data than they're analyzing, and that's your data. And you can then upload it to these kind of third-party sites like Jedmatch, which sort of is designed to allow folks in the ancestry space to compare their DNA to folks in the 23-and-me space, right?

Because normally those two groups don't talk to one another. So does one type of website offer more privacy than the other? Well, it depends what you mean by privacy. So 23 and Me and Ancestry DNA and other services like those are more like a walled garden. And so they, you know, will inform their users that they may be required to comply with court orders, but that they will do their best to fight them. Whereas the GEDmatch platform from what I've learned about the site in the last couple of days is really a much more open system, right?

It doesn't make as many promises. It's not a for-profit enterprise. Amy McGuire, before this case, you know, people were not talking much about the bioethics of something like this happening where, you know, people could submit their genetic information and not realize what could be done with it? What were the biggest concerns about genetic privacy before then, and what concerns do you have now? So I think genetic privacy has always been a concern regarding genetic testing, genetics research, and since the beginning of these sort of direct-to-consumer genetic testing companies coming on to the scene,

there's been a lot of concern about privacy policies and making sure that people's information is protected. And always there's going to be a trade-off. And so the people who are using these services tend to think that the benefits to them, whether it's finding out information about potential health risks through companies like 23M or whether it's connecting with potential genetic relatives, finding out more information about their ancestry, those are considered benefits that people have typically felt when they use. these services outweigh the privacy risks, although there's a big concern that people may not fully understand what risks

they're going to be encountering when they do these services. But it seems now I was thinking about this. This is almost a parallel to the Facebook scandal that's going on now. Hey, I didn't know you were going to do this with my information.

Now we're hearing something like, hey, I didn't know you're going to do this with my information about my genes? Yeah, I think that's a very good point. I think that in the big data world that we live in now, it's not just genetic information. It's social media information. It's information that we share online. It's information that we, you know, this has come up with regard to people's cell phone activities. So it's really, we're kind of at an inflection point, I think, in our society and in our legal system to be thinking

about how do we handle privacy issues in the context of big data where you're inevitably going to be sharing your information, whether intentionally or unintentionally with other third parties. Our number 8447248255. You can also tweet us at Cy Fry. Amy, it seems like there's really no way to escape having my genetic information shared online, even if my third cousin takes the test and uploads it somewhere. That means I'm related to my third cousin, and so something about me is up online, is it not? Yeah, I think that's true. And, you know, we were, I participated in a study that was done out of MIT several years ago,

where we sort of first looked at this familial matching through genetic databases. And it was surprisingly easy, actually, to identify people through their familial matches on genetic genealogy databases. And one of the things that struck me is you may not even know you're related to the person who is, has uploaded their data. They could be, you know, fourth, fifth, sixth generation cousins who you had no idea, you had any, you don't know who they are and you no idea you had any relation to them, and yet their information could in some ways be used to identify you. So I do think it raises really important questions, particularly with, that's what makes the genetic piece of this slightly different than, like, Facebook information, because genetic

information is in and of itself inherently familial, and so you can tell information about your genetic relatives through your own DNA. And Natalie, could a genetic testing company refuse to share my information with law enforcement? Are they protected under any sort of law? So police will probably start with a request, and these kinds of companies can refuse a request. Their ability to refuse a subpoena or a judicial warrant is much more limited. So as happened in this investigation, there was a subpoena made to a different website not to Jedmatch, a subpoena requiring that company to divulge.

a particular user's name and payment information to allow investigators to continue on in their investigation on that family tree. And generally speaking, once you have voluntarily shared your data with a third party, like a genetic testing service, the Fourth Amendment's protection against unreasonable searches and seizures against warrantless searches and seizures simply doesn't apply. That's the current law. Now, the Supreme Court is currently considering a case that may, may, right? We don't know how it will be decided yet, but may reshape this idea that once you share your data with any third party, however nominally, the Fourth Amendment, you know, kind of now that data is outside the bounds of the Fourth Amendment's warrant requirement.

So this could really change, but it's unclear whether or not it will. any data or just genetic testing data that cases before the court? Oh, so the case before the Supreme Court has to do with your cell phone location data, the historical cell phone location data. So that case isn't about genetic information, but what the Supreme Court

says about the Fourth Amendment's application to that data could help us reframe how we think about privacy and other contexts, including for genetic information. It's also worth noting that there are a number of federal

laws that protect data of certain kinds in certain settings. So there is the HIPAA privacy rule, which protects health-related information. But it's not at all clear that the kind of data we're talking about when we're talking about these direct-to-consumer companies or these genealogical data platforms, right, are dealing in health-related information. And there's other kinds of protections that apply to scientific research data, but that's also not really applicable here. So we're kind of in a space without a lot of legal protections against a subpoena or certainly a warrant and maybe even a request. We're going to take a break, and when we come back, we're going to talk more with Amy McGuire and

Natalie Ram, and our questions from 0844-8255. You can also tweet us at the SciFri. Stay with us. We'll be right back after this break. This is Science Friday. I'm Ira Flato. We're talking this hour about privacy issues, genetic testing, and law with my guest, Dr. Amy McGuire, Professor of Biomedical Ethics, Director of the Center for Medical Ethics and Health Policy at Baylor College of Medicine. Not only Ram, assistant professor of law at the University of Baltimore Law School, our number 844-724-8255. You can also tweet us at Cy Fry. With the Golden State Killer case in mind, you know, I've been following, we have a lot of listeners in California, the case, the cold case of the Zodiac Killer. And people are talking about, hey, maybe we could find the Zodiac Killer using DNA evidence. Have we learned anything from the Golden State Killer case, Amy and Natalie, about going after the Zodiac Killer? Is there any lesson there?

At a minimum, it shows that this kind of technique can produce results, right? And that can be a big win for public safety. I have some concerns about the use of genealogical databases and other kinds of genetic databases in this way, in part because they don't fall within the scope of traditional legislative structures. for whose DNA should be searchable by the police. Do we, you know, one of the things that we've been noticing more and more as law enforcement gets more active is that we've had facial recognition. We have thousands and thousands of cameras around big cities and other places, just constantly taking pictures of people's faces or in public places, airports, whatever. Could we expect now to see a collection of a database? of DNA also randomly taken and used?

So I think many people would like to see that happen, I think both from a public health perspective but also from a research perspective. The National Institutes of Health has talked for a long time about having a national database of genetic information for research purposes. And we've really gotten a lot of pushback. I think our fundamental values in the United States around privacy and concern. about genetic discrimination and how people might use genetic information still are very pervasive. We do have all 50 states and the federal government sort of signed on to this idea of having a database of genetic information from

suspected and convicted criminals to varying degrees. And there seems to be general public support for that. Although, as you start to move a little bit further out into people who have been suspected of nonviolent crimes and things like, that, I think there's less support for them being included in these forensic databases. So expanding out from that to a non-phornsic database where we just collect people's DNA at birth, for example, and just have a national registry, I think it would be, I would be surprised if we had widespread public support for that in our country. Other countries might have an easier time with that. Okay, let's go to the phones.

8447-24-8255. Let's go to Cleveland. Hi, welcome to Science Friday. Yes, this is a Bill Eaton calling from Cleveland. I was thinking of some way we might be able to get around this by either going through doctor-patient privilege or attorney-client privilege. What about that? Could you deny, you know, saying you can't give that to me?

You're doctor or a lawyer. So with the data... Go ahead, Amy. I'm sorry. I was just going to say, so with the... The genetic genealogy databases and even the direct-to-consumer genetic testing companies, they go to great lengths to not establish a physician-patient relationship with their customers and to basically say that the information that they're giving back is not medical information.

It's not diagnostic information. And so oftentimes they don't even have a licensed health care provider involved in their services. So in those circumstances, there is no physician-patient-patient-patient-patient relationship or even attorney-client relationship that would privilege the information in the way that you would if it was your medical information and it was obtained during the course of your medical care. Let's go to the phones to Kit in Sausalito, California. Hi, Kit. Hi. I had kind of a similar question. I think your panelists began to discuss it, but hasn't the privacy of genetic information been previously discussed in terms of genetic discrimination?

finding some family or person who's predisposed to a disease would give insurance companies a leg up? Good question. There is a federal law, the Genetic Information Non-Discrimination Act, which prohibits employers and health insurers from using genetic data to make decisions, employment and insurability decisions. But that's as far as that statute goes. It doesn't apply to law enforcement. It doesn't even apply to law enforcement. and it doesn't even apply to life insurance underwriters. So the federal law on your genetic privacy is pretty limited. And again, though there is this federal privacy rule for health-related data, as Amy pointed out before, the kinds of services that we're talking about here with these direct-to-consumer companies,

they're disclaiming that what they're providing is health information. A tweet from Shane who says, I'm concerned that the path, this is for, following will interfere with people trying to reunite with biological families like adopted children looking for their long-loss families. Shouldn't some consideration be given for adopted people? Yeah, so I'll just say I think that that raises a really interesting and important issue that's kind of tangential to what we're talking about today, but these genetic genealogy websites and these direct-to-consumer genetic testing company websites can be used for a whole variety

of purposes with regard to identifying genetic relatives. And that's one way in which they're frequently used for children who have been adopted to find out more information and connect with potential relatives. I also have heard of many anecdotal cases where obviously there's the risk of non-paternity being identified. So you find out who you thought was your biological father, was not your biological father, through these, or you find out somebody else was your biological father, and also cases of anonymous sperm donation and being able to identify the anonymous sperm donor, which kind of undermines

that whole system. So I think there's a lot of ethical and legal questions about how these databases are being used in ways that kind of conflict with our existing expectations and laws around anonymity in certain circumstances. Is there any way to sort of regulate the industry? You know, people always say, oh, you're going to regulate everything, but should there be a little more regulation here? The United States has often taken a hands-off approach when it comes to reproductive technologies and genetic technologies and the like. So that said, you know, every state and the federal government has set out whose DNA should be collected and stored in a

database for law enforcement use to solve crime, right? And right now, the Supreme Court has previously signed off on the collection of genetic information from people convicted of crimes or arrested, but not yet convicted of crimes. No state has suggested that everyone should go into a database. And for reasons that Amy identified earlier, I think that would be unlikely to gain traction And so one of the things that troubles me about the use of genealogical databases is that once you start using these non-phorxic databases, you know, the net grows ever wider. And as far as I can tell, only a limited, it's only a matter of time before we're all in. Say that again, it's only a matter of time. It's only a matter.

It's only a matter. It's only a matter. Yeah, well, absent some strong legal regulation, right? And it, you know, Congress or a particular state could enact, you know, some sort of genetic privacy law that says in law enforcement can't get your, can't use your DNA just because it's, you know, shared with some third party like a 23 and me, jeddmatch.com. A state or the federal government could do that. Absent that kind of action. I think what this case demonstrates is that when investigators can't solve the crime using the existing mechanisms and tools they have at their disposal, they look for other sources, new ways to expand the net of whose DNA they can search. All right. And on that chilling note, that's all the time we have. We will revisit this. I'm sure Dr. Amy McGuire, Professor of Biomedical Ethics and Director of the Center for Medical Ethics and Health Policy. at Baylor College of Medicine and Natalie Ram, Assistant, Professor of Law, University of Baltimore School of Law.

Thank you both for taking time to be with us, and have a good weekend. Thank you. Most dog people will agree your dog is one of the most valued members of the family, right? I have four grand dogs, and I can tell you that we love them all. And it seems to go both ways when you get, you know, when you're excitedly greeted at the door, don't you say, hey, the dog knows I was gone? it's happy to see me. Or they tilt their head from side to side when you're talking.

It seems like they're trying to understand you. But how true is this? What is your dog really thinking or feeling? How does your prized pooch make sense of the world? These are exactly the questions my next guest had about his own dog. And he's been testing out those ideas by training dogs to sit quietly for a long time in an FMRI scale. and looking at how their brains process information. Gregory Burns is Professor of Psychology at Emory University.

Thanks for joining us, Dr. Burns. Hey, it's great to be here. And I want to let our listeners know that our latest macroscope video features Dr. Burns' lab where you can see all of these dogs hopping into the scanner. It's up at our website at ScienceFriday.com slash dog brains. It's just a wonderful video. We get to see how you work with these dogs. and the first thing that surprised me is how do you get them to sit so quietly inside the MRI machine?

Lots of treats and lots of fun. Seriously, the whole way that we train the dogs is built around this idea that the dog doesn't know what an MRI scanner is. So they don't have any reason to be afraid of it, not knowing what it is. So we just train the dogs to associate all the aspects of it, the noise, the steps, the patient table, the thing we call the head coil that they have to put their head in. We train them outside the scanner typically for a couple of months, and the idea is to make it a game. So everything that we do is paired with either food, if the dog is really food-motivated or with play. And so to them, we hope it really is just a game.

And if you have questions about your dog or you want to express how you think really dog feels about, how do you dog feels about, draw a number, 844, 724-7-24-8-4-4-4-4-4-4-4-4-4. 8255. You can also tweet us at SciFRI. Now, you are a neuroscientist. Why are you interested in studying dog brains? What scientifically interests you besides have a dog, which could be a reason enough, I think? Well, I think that is actually a reason enough. And, you know, my training is as a neuroscientist, I spent 20 years studying the human brain, specifically how human brains make decisions and what motivates us. And the dog project really was just a crazy idea initially, which was, hey, can I train my dog to go in the scanner so I can see what she's thinking? And initially, it was just an idea in search of a question, and then it grew from that.

And I think we've backed into a lot of the science now in terms of the things that are important about dogs. And so I would say dogs are special, and that's not just because I'm a dog person. dogs are the first animals to live with humans, and so they represent a key link to our past, you know, what makes us human in terms of how they see us, you know, how do they respond to us? What is it exactly that they're doing that allows them to live with us so easily

as well as other animals? These are very basic questions that, you know, we don't fully understand, and yet they also tell us something about being human as well. I'm Ira Flato. This is Science Friday from WNYC. studios. Talking with Dr. Gregory Burns, professor of psychology at Emory, also a happy dog owner, and someone who's putting dogs through his MRI. Most past studies about how dogs think are behavioral

studies. You're actually looking at brain scans, right? The dogs are between being shown objects. That's sort of a bit of a new idea. What can you get from a brain scan compared to a behavior your study? Yeah, that's a great question. And so when I started doing this, you know, we started with doing behavioral experiments with dogs and just asking them to make choices. So kind of a classic example that I think everyone wants to know the answer to is, you know, does my dog love me? Well, that's a really hard question to answer. So some of the ways you might go about doing that with a behavioral test would be maybe, you know, you put a person in one part of a room and a treat in the other. And you might see.

which does the dog choose. Well, we've done that sort of experiment, and I'll tell you, the dog's strategies on these things kind of run the gamut. You know, some dogs just pick one side and they just go to the same side. They just go left constantly. Or some dogs will sample back and forth,

or some dogs will just stick with a thing that they like the most or paid off the first time. Or usually they just look for the nearest human and see what cues they're giving them. So it's really difficult, I think, to understand what's going on inside a dog's head just from their behavior. And it's very challenging to design good behavioral experiments because dogs are so attuned to the humans around them that they suffer from what we call the Clever Hans effect. That's interesting. The who?

The Clever Hans was a horse in the 20s who supposedly could count or do arithmetic. and what the horse was actually doing was picking up subconscious cues from their handler. So in psychology, it's called the Cleverhans effect when the animals pick up cues from the people around them. So it wasn't Mr. Ed. No. So have you, by looking at scans, can you solve that problem? Does my dog love me?

I think we can. I think we're getting close to it because when we're scanning the dogs, they're completely awake. And they're not really doing anything in terms of a task. We're not generally asking them to make choices. They're just in the scanner the same way a human would be. And then we just show them things. So in terms of, you know, dog love or food, you know, we've done that type of experiment where, you know, we present a cue to them that signals, hey, you're going to get a piece

of food in a few seconds. And we register the brain's response to that cue signaling that. And then we do the same thing with a different cue that says, hey, your owner's going to pop into view and praise you and say, good girl. And all we do is we look at specific circuits in the dog's brain associated with reward and see which of those cues is most effective at activating it. And I think most people will be happy to know that, at least in the dogs that we've studied, we found that both were equally activating to the dog's reward system,

meaning that they do like food, obviously, but they also like their owner or they're human just for the social contact itself. I'm glad you could verify that when my dog is jumping up and down to see me, that it's really loved. So we're going to take a break and come back and talk more with Dr. Gregory Burns, Professor of Psychology at Emory University in Atlanta. Our number 844-24-8255, lots of dog lovers out there. And we'll talk more, take your phone calls,

and find out whether Dr. Burns wants to throw a bone to the cat lovers out there to see what they're thinking or is that. That's a whole different story. We'll find out why after the break. Stay with us. This is Science Friday. I'm Ira Flato talking with Gregory Burns, Professor of Psychology at Emory University in Atlanta.

Dr. Burns, do people call you the dog whisperer at all? Only behind my back. As you can imagine, we have so many questions. But let me get, because we have, I'm sure, a lot of cat-less, cat owners, out there. Can you do the same thing with a cat or do you not care to or any different results? You know, I talk a lot and when I give talks, inevitably someone comes up to me and says, I have a cat who can do this. And I say, okay, great, show me. And I haven't met that cat yet.

I just want to tell other people, just remind people that we have a great macroscope video up there on our website in Dr. Burns' lab where you can see all of these different dogs hopping into the scanner. It's on our website at ScienceFriada.com slash dog brains. Lots of lots of people with questions. Let's go to them. Let's go to Mary Grace in Birmingham, Alabama. Hi, Mary. How are you today? Hey, fine. Go ahead. Welcome. Awesome. Thank you so much. So Dr. Burns, I was wondering if you could speak a little bit about what you've seen regarding different personalities in dogs. My question is brought on because I honestly think my husband thinks I'm a little crazy, but we're.

We have two very large dogs, both with very distinct and separate personalities. And in fact, one of our dogs, one of the main differences in my mind is how one of our dogs reacts when we leave and then come home. If we're gone for, you know, an hour or two, one of our dogs who walk in the door and he's like, hey, how's it going? Glad you're back. But the other dog has a complete breakdown. It's like this cataclysmic event in his mind that, oh, my gosh, I'm so glad you were back.

And he'll cry, he'll fall down on the floor, literally crying and whining and shaking. for about 15 minutes. So I just think it's so interesting how you can have two dogs that grew up in the same household for their whole lives, just like you can have two kids, who have such drastically different responses to the same event. Can you talk at all about what you've seen regarding personality differences between dogs and your studies? Sure. And thank you for that question. That is a really good question. And, you know, the issue of personality is something very dear to my heart and one of the growing issues. And you know, the issue of personality is something very dear to my heart. And one of the growing areas in this line of research. So when we started out, which was now seven years ago,

we were kind of pursuing how the dog brain just works. And now the questions are really about what makes an individual dog's brain work and why is that dog's brain different than another dog's brain? And so some of the clues that we've had to this comes from correlating these brain responses with aspects of the dog's personality. Now it's sometimes difficult, to measure dog personality, but there are various ways you can do it. So you can go online, you can find there's one questionnaire that's called the C-B-A-R-Q or the C-B-A-R-Q made by the folks at Penn, and you can answer about 50 questions about your dog, and you'll get a personality and temperament profile.

So we've actually used that in some of the dogs that are trained for the MRI. And what we have found is that there are specific personality types that show up differently in their brain responses. The one that probably makes the most difference is how aggressive the dog is. You know, are they aggressive to other dogs? Are they aggressive to other people? And what we found in those dogs is there's a particular part of their brain called the amygdala, which is associated with what we call physiological arousal.

It's the part of the brain that really gets you revved up. And those dogs seem to have more of a hair trigger on that structure, particularly when they encounter a person they haven't met before or sometimes perhaps even a dog they haven't met before. So there are definitely differences. So I would say you're not imagining this and you're not crazy. Yeah, because I have seen very great. I've seen the same thing that you have with dogs.

When one owner goes away, the dog is just whining until the owner comes back and just jumping up and down and is happy to see. Dr. Gregg, Dr. Burns, is that a measurable kind of thing? I think it is. I mean, it's very messy to measure these things kind of when it's happening. And that's why I think we can get clues to what's going on inside their minds by using the imaging. Because in many cases, when they're in the scanner, they're in kind of a more contained state. And yet what's amazing to me is even though the dogs are still in the scanner, they're all doing what they're supposed to do,

we can still detect differences in how they're reacting to these very salient cues. So they are experiencing these things. and I would call them emotions, actually. Can you test and have you tested for, you know, I've watched dogs, watch other dogs on television? Do they know that that's a real dog on TV? Oh, that's a great question, and I would love to know the answer to that. We've done some work with human faces,

so we've projected human faces and dog faces on a screen while the dog is in the scanner. And what we found is that just like humans, the dogs have a part of their brain and their visual systems, that seems dedicated to processing faces, both human and dog faces. So we know that they have the hardware there to process these things. Now, where it gets confusing, I think, is the 2D aspect to it. So humans are very natural. We're very good at just abstracting these 2D images,

and we know that what we see on the screen represents the real thing. We're presently trying to figure out whether dogs can do that as well, or is it just kind of low-level features of the images? Here's a tweet that I've always wondered about from Dawn who says, how do dogs process and learn our language? India, I think, is her dog, can seem to understand a conversation not directed at him. Yeah, this is amazing. These are great questions.

So the thing about language that amazes me is that, you know, we humans talk constantly, frequently without saying much, too. And it's amazing that dogs can kind of pick out salient words out of this kind of stream of gibberish coming from us and somehow magically know what it means. So the question, and we're pursuing this right now, is does a word to a dog mean the same thing as the word to a human? And this is a very complicated cognitive skill. So we know that a word, if I say stick, for example, we call up a mental image of a stick and we know that that word is, a symbolic representation for the real thing. Now, it may be that dogs don't quite have all of that neural hardware to do that.

They may be able to associate the sounds, perhaps, with the object, but we're not sure that they have kind of the ability to abstractly represent it, what we call semantics. But we're trying to figure that out. Yeah, you know, they know the words. You have to spell W-A-L-K instead of say it because we're going to hear that word in the middle of a conversation. Exactly.

Exactly. Pick it up and do that kind. Let's go to Jacksonville, Florida. Hi, welcome to Science Friday. Hi. I was wondering what the potential was for treating the psychological traumas of dogs that are rescues of abusive homes. Yeah. Can you see that in Brainstead?

That's a great question because so many dogs come from rescue shelters and they have been abused. And I know many of those dogs. and they really are hurting a lot of times. Yeah, that's a great question. And so I think kind of the deep question here is, you know, what kinds of memories do dogs have? Do they remember their past and kind of what, you know, kind of traces does that leave in their brain and their personalities?

So these are very difficult questions to answer, even in humans when you go looking for them in the brain. We're just beginning to scratch the surface. So part of the difficulty in doing this sort of thing with dogs is the dog has to be calm enough and trainable enough to go in the scanner in the first place and not be so anxious that they would startle and run away from the noise. If we can get them over that hurdle, if we can train them to do that, then we can begin to answer these questions. As we've grown our project here in Atlanta, we've trained now close to 50 dogs, many of whom do come from shelters. Many of them are adopted dogs. And it's a great question.

I haven't yet looked at the differences between, say, a rescue dog and a purebred dog that came straight, you know, from a breeder in kind of a very nice environment. Seems like a lot of work for one guy to do here. No, I mean, you know, when you study basic research, you really need more than one researcher working on stuff, more than one center. There are others groups. There's a couple other groups around the world. that are starting to pick this up. And, you know, I've gone around proselytizing the benefits of doing this kind of research. The area actually that I'm really passionate about right now is actually the benefits for the dogs themselves.

So people don't really think about this when we're doing this research, but we've been doing this so long that the dogs are aging out of our project. Because of the short lifespan of dogs, we are picking up health. problems associated with dogs like cancer. And because they're trained to go in the MRI, we can image them so quickly that we're in a position now to help the dog's health pick up cancers and other health problems before the dogs show symptoms, or at least before they tell their humans that something's wrong. And who's paying for this kind of research? So the research is primarily funded by the DOD, Department of Defense, specifically the Office of Naval research. Also, DARPA has funded this work. And of course, the follow-up question is, why do they care?

You saved me. You saved me. The simple answer is that dogs, apart from being companion animals, play important working roles in a variety of settings, from service dogs, assistance dogs, to police dogs, to military working dogs, in all these roles. And it's, it's, it's critical, especially in roles where people's lives depend on them, to understand what the dog is seeing in their world and how can we be better attuned to them doing their job and working with humans. Here's a tweet from Pamela Kennedy who says, do dogs have a good concept of time passing? When we go on vacation and put them in boarding, do three days versus seven days make a difference to them? Yeah, great question. It's not a question that we can really answer with brain imaging.

But just kind of common sense, the answer is yes, they obviously have a sense of time. And the reason I say that is because all animals have some kind of sense of time, and they can do it in a variety of ways. They have internal rhythms, circadian rhythms that govern just kind of the 24-hour cycle. You can gauge time by how hungry you are, by whether your bladder is filling up, by the change in lighting outside. So these are all cues that dogs pick up on. Interesting.

Let's go to the phones to Houston. Gabby. Welcome to Science Friday. Hi. Welcome. Thank you for taking my call. Hi. Go ahead.

Okay, so I don't know if I had a good question and comment, but my dog, I've had her for like four years, and her name is Lexus. She's a Rhodesian Rich back, and I was wondering if, because we were talking about the words, and if they pick up on words, my dog literally knows what it's side-by-side, and you should give me a kiss. Like when I'm coming to the house, she's like, unlike any other dog I've ever seen, her personality is on point with mine. I mean, like, I could just look at her and tell her, you know, it's something, and she'll do it. So when you were talking about treats versus you, I hands down 100% know that she'll

shoot she pick me. So I was wondering if you, what dogs were you testing? And if that has to play in anything, like, like, the doubt that you have any of your data. I risked bulls on my life as well, and these creatures are own, like, emotional creatures as well. So, I mean, what would that be? Yeah. Now, let me rephrase it. Thanks for calling.

This is Science Friday from WNIC Studios. Yeah, it's getting a little muddy there, but the phone call was degrading a bit. But what you wanted to know is that the dogs differ by breeds in how they perceive you and react to you. Yeah. Well, it's a tough question because, you know, even though we've. We've studied, you know, 50 dogs in Atlanta, another 50 in California. There's over 100 breeds, so we don't, you know, we don't have the numbers to answer that question. But what I can tell you when we did one study that we partnered with actually with Canine Companions for Independence to study service dogs,

and those dogs were all golden retrievers and Labrador retrievers and mixes. And so even within one very homogeneous breed, we find tremendous variation. And this gets back to the previous question about personalities, that we find as much variation in the dog's brain responses as we do in the equivalent experiments in humans. So dogs are as different from each other as humans are different from each other. Now, what that means for breeds, it's probably even greater between the breeds. We have a tweet in from David.

It says, I'm very sure that the dogs don't pass the classic mirror test. yet they are not that site priority-minded. Anyone done any scent-based testing of recognition, self-recognition in dogs? Yeah, that's a great question about self-recognition. So there have been a couple of groups who have studied this. So one study, actually the original study looked at, do dogs recognize their own urine?

And the answer appears to be yes. Although that's not quite the version of the mirror test. So another study recently replicated that, by seeing if dogs recognize their urine adulterated with some other substance. And the answer could be yes. But the issue of self-awareness, I think, is fascinating. And just because dogs don't recognize themselves in the mirror, generally does not mean that they're not self-aware.

It may be, like you just said, they're not visually self-aware. So what is the one mega question that you would like to answer if you could? I think the mega question is, comes back to the language one. And this is something that I'm deeply interested in, which is what do dogs get out of human speech? What are they processing? How do they do it? And do they have a mental imagery in their heads that, you know, translates from words? And that's, you know, that's where we're currently working.

And how do you do that? How do you test that out? So there are several ways. It's obviously saying language in humans as complex. It's an order of magnitude difficult in dogs, but the idea is that you know, you teach dogs the names of things. So we're currently working on an experiment where we've taught the dog's names of objects. And then you have the people speak those words to the dogs while we're skating their brains and see if we can tell what kind of representation that has. Does saying the word ball or stick cause some kind of visual activity as if they're visually imagining it?

Can you do they? Can they? I can't tell yet. Not yet. It's difficult to say. Watch this space. Exactly. Well, we'll have you back when we can learn some more stuff.

Very interesting, Dr. Burns, or Gregory Burns, Professor of Psychology at Emory University, and you can watch our latest macroscope video of Dr. Burns in his lab and all the dogs sitting and staying in the MRI machines on our website at ScienceFriady.com. Thank you for taking time to be with us today. My pleasure. One last thing before we go. Science Friday's hitting the road to Pittsburgh. just a few weeks, taking the stage at the Carnegie Library Music Hall, Saturday, May 19th.

We'll have roboticists and artificial intelligence designers, chairs that build themselves. Wow, smart clothing that senses your skin to keep you cool. Join us tickets and info at ScienceFriady.com slash Pittsburgh. ScienceFriety.com slash Pittsburgh. Pittsburgh at May 19th, the Carnegie Library Music Hall, ScienceFrily.com slash Pittsburgh. Jay Leiderman composed our theme music, and of course you can find us on all social media or up there all week long. So, like, really, every day is Science Friday. Have a great weekend.

Happy Cinco de Mayo. I'm Ira Flato in New York.