NEAL CONAN, HOST:
This is TALK OF THE NATION. I'm Neal Conan in Washington. In the coming months, the Supreme Court is expected to rule on two cases involving police dogs and the limits of reasonable search. Does the trained nose of a detector dog provide probable cause, and does a sniff from a front porch constitute an illegal search?
Both cases also raise questions about just how reliable these dogs, how they're trained and how they're used. If you work with dogs in law enforcement, call and tell us what works, what doesn't. 800-989-8255 is our phone number, the email address talk@npr.org. You can also join the conversation on our website. That's at npr.org. Click on TALK OF THE NATION.
Later in the program, we begin our Oscar Doc series with "Searching for Sugar Man." But first detector dogs. David Latimer is chief of police in Harpursville, Alabama, and runs FSI Canine Academy. He joins us by smartphone from his office Harpursville. Nice to have you with us. David, you there?
Well, we can hear him; apparently he can't hear us. So we'll try to contact him again. He's the chief of police in Harpursville, Alabama. He conducts seminars and training for dogs and their handlers across the country. David Latimer, are you there now?
DAVID LATIMER: Hello?
CONAN: Hi, you're on the air.
LATIMER: Oh, thank you.
CONAN: Welcome to the program. You not only conduct this training, you work with dogs in your day job. You have a dog you walk on the beat with.
LATIMER: Well, I don't do a lot of beat walking anymore. I'm a chief now. I have done a lot of beat walking.
CONAN: And what is it like when the dog's with you on a patrol like that? What goes through your head?
LATIMER: Well, it's like having a friend. I mean, especially a friend that can't talk back.
(LAUGHTER)
CONAN: And a friend who will always be loyal.
LATIMER: Will always be there and forgive you no matter what kind of mistake you make.
CONAN: And tell us a little bit about training of detector dogs. Are they trained for specific things?
LATIMER: They are trained for specific things, everything from detection of bedbugs to diabetes, the onset of diabetes, seizures or comas, narcotics, explosives, just about - you name it. If it's unique, has a unique odor, a dog can be trained to find it.
CONAN: And I've read that they can even be trained to discover cell phones in prison cells, that sort of thing.
LATIMER: That is correct, yes.
CONAN: What is the distinctive odor of a cell phone?
LATIMER: I don't know.
(LAUGHTER)
LATIMER: The dog knows, but I have no idea. But, you know, we don't know what the specific odor or specific constituent is in cocaine that a dog detects. We just know they can do it.
CONAN: They can do it. So obviously some things must be easier to detect than others.
LATIMER: Yes, absolutely. Some materials are more - produce more vapors or particulates than others. Gasoline, for example, for what's known as an arson dog is a fairly easy substance to detect because it's so volatile.
CONAN: And the dogs you worked with when you were on the beat, what were they trained to do?
LATIMER: Well, I worked with dogs on explosives, narcotics, evidence, cadaver search dogs.
CONAN: So that they would look for buried bodies.
LATIMER: Yes, sir. Well, not just buried bodies, evidence of a body, body fluids, body parts, not necessarily a whole body.
CONAN: And how does a dog signal to you that they've found something?
LATIMER: All my dogs are trained to perform what we call passive alert, in other words they sit when they detect an odor and then point directly at the source of that odor with their nose.
CONAN: And obviously the handler has to be trained, as well, I gather in part not to transmit to the dog some hints about what they might be looking for.
LATIMER: That's correct. They - that among several other things, but it is - the handler's training is just as important, just as indispensible as the dog's training, that's correct.
CONAN: And are these dogs always on leashes?
LATIMER: I'm sorry, are they what?
CONAN: Always on leashes?
LATIMER: No, no for example my search dog, I rarely ever put him on a leash, cadaver dogs the same way. I allow them to do what we call a free search. I just want them to follow their nose and take me to the odor of interest.
CONAN: And as some dogs better at this than other dogs? Are some dogs better at some things than other dogs?
LATIMER: Well, there's a perception out there that they are, but it's more of a misconception or a prejudicial view, if you will, that bloodhounds are the top-notch tracking dogs, or German shepherds are the top-notch narcotics dogs. Most any dog has the olfactory sense to make it capable of detecting the odors that we typically need to find.
But, you know, an animal that's great at detection, by the way, is a pig, but can you imagine a police officer leading a pig around a car?
CONAN: It might lead to a few unhappy comments.
LATIMER: Yeah.
CONAN: So where do you get these dogs?
LATIMER: We import dogs, but we get a lot of our dogs, especially for the private interests, we get them from shelters. They are just, it's really sad how many dogs, great dogs are available in shelters that are just kind of are throwaways, so to speak.
CONAN: And let's get another voice into the conversation, David Latimer, police chief and dog trainer Lawrence Myers, an associate professor of animal behavior and sensory physiology and medicine at Auburn University. He's conducted research on detector dogs for more than 30 years and joins us by Skype from Auburn. Nice to have you with us today.
LAWRENCE MYERS: Thank you very much.
CONAN: And what does your research show in terms of how accurate and reliable detector dogs are?
MYERS: That's a really general question that has a number of specific answers, I'm afraid. Some dogs can be extremely accurate and reliable while others have been proven not to be so at all. There's an enormous variability there.
CONAN: So it varies by individual, by breed, what?
MYERS: Not so much by breed, but - and I don't have enough numbers to give you a good answer on - in terms of the influence of the individual. We do know that type of training and type of maintenance training and health of the dog, all of these certainly have a significant impact on the reliability and accuracy.
CONAN: So the quality of the training obviously has an influence.
MYERS: Yes, it does.
CONAN: And so is there, for example, any good numbers on how often a narcotics dog, for example, comes up with a false positive?
MYERS: Not really. The numbers that are usually floated around deal with the number of alerts that a dog has out in the field versus how many times that alert resulted in finding actual narcotics. There's an - pardon me, an excuse made often when no narcotics are found that, well, narcotics mist have been there, and it's a lot of the handlers and trainers call residual odor. The chemicals that do remain after a narcotic has been present.
Now sometimes that true. You can demonstrate that. Now I can put a narcotic in a box for about a half an hour, take it away, and a trained dog will still come around and alert on that. So yes, that does happen. But how frequently it happens, that's the big unknown. So really it's the sort of thing that has to be studied in controlled conditions, and the numbers there are harder to get.
CONAN: And it is safe to say, though, dogs are hardly infallible.
MYERS: They are definitely not infallible. Even very good dogs make mistakes.
CONAN: Let's get some callers in on the conversation. We want to hear from those of you who have worked with dogs in law enforcement. What works? What doesn't? 800-989-8255. Email talk@npr.org. And Jim's(ph) on the line with us from Arcadia in California.
JIM: Yes, thank you for taking my call.
CONAN: Go ahead, please.
JIM: I've worked with a canine handler in law enforcement in training his dog and maintenance training with his dog. And one of the things that I've learned is there are dumb dogs just like there are dumb people, and there are intelligent dogs just like there are intelligent people. I mean, a lot of it has to do with training, but the quality of the training varies.
We were in a - I was working him as a city beat cop with his training, and we went to a competition with a bunch of state police, and his dog completely smoked all the state police dogs. I mean, he absolutely just walked all over these other law enforcement agents with their dogs, and...
CONAN: What was the competition like? Tell us a little bit about it.
JIM: Well, there was a tracking competition, there was a - there was a narcotics, or drugs, if you will. There was explosives. There was an obstacle course. There was just basically crowd control, obedience during crowd control, all these different competitions. And it was just absolutely amazing to watch this dog that I had worked with and been a part of.
I wasn't the actual trainer, but I was helping and just, you know, absolutely just smoke everybody in the competition on this. You know, it was basically supposed to be a display for government officials and for upper law enforcement, and it was really incredible to see the variance and the difference in abilities of supposedly professional handlers and their animals.
CONAN: David Latimer, does that make sense to you?
LATIMER: It does. I only disagree with one thing Jim said, about some dogs being dumb. Dogs know basically what they're taught, and a dog's - dogs are just like people. They have different levels of drives an initiatives, so to speak. And typically when I found a dog that was imputed as being dumb, it turned out that dog really understand what was being required of him.
But I have seen the kind of situations that Jim describes, where somebody kind of comes out of nowhere, so to speak, and smokes a lot of the supposed professional dogs.
CONAN: Jim, do you still work with dogs?
JIM: No, well, I work with dogs on an informal basis now. I'll train my friends' dogs with obedience and various things, but, you know, not so much the law enforcement part of it anymore.
CONAN: Well, thanks very much for the phone call, appreciate it. That was interesting.
JIM: Thank you.
CONAN: After a short break, we'll hear a little bit more about two detector dogs with cases before the high court, Franky and Aldo. If you work with dogs in law enforcement, call and tell us what works, what doesn't, 800-989-8255. Email us, talk@npr.org. We'll be back in a minute. Stay with us. I'm Neal Conan. It's the TALK OF THE NATION from NPR News.
(SOUNDBITE OF MUSIC)
CONAN: This is TALK OF THE NATION from NPR News. I'm Neal Conan. With two cases before the Supreme Court featuring the work of drug-sniffing dogs, we thought we'd talk today about the animals themselves; how they're selected, how they're trained and the legal limits on the work they do. Specifically, though, the two dogs in question are Franky and Aldo.
Franky the dog who was walked up to the front porch of a private residence is a kind of rockstar dog. He's a chocolate lab. His record for drug detection is near perfect. Aldo, ah, that's another story. He's a German shepherd who sniffed out ingredients used to make meth stashed in a truck outside Tallahassee, but the Florida Supreme Court threw out the conviction based on his find.
Aldo did not have a record of reliable contraband detection. So his alert was suspect. If you're in law enforcement and work with dogs, call and tell us what works, what doesn't, 800-989-8255. Email us, talk@npr.org. You can also join the conversation on our website. That's at npr.org. Click on TALK OF THE NATION.
Police chief and dog trainer David Latimer and Lawrence Myers of Auburn University are our guests. And Larry Myers, can I turned to you? In your research, do some breeds do better than others?
MYERS: Actually, if I just look at sense of smell, which is something I did for some years, measuring how well they smell - and I probably tested upwards of 4,000 dogs total - there was so much variability within a breed. So we have one German shepherd that has a fantastically acute sense of smell, and we have another that, well, it...
(LAUGHTER)
MYERS: ...it gets laughable. They vary all over the place within a breed. So you really can't pick a breed, at least in my experience and my research.
CONAN: And have you found in your research that some methods of training work better than others?
MYERS: I am hesitate to state anything very definitively there. But what's being shown so far is that there are many ways that you can effectively train an animal. Some of the really critical things seem to be how you test it and how you maintain that training.
For instance, there's two situations, one is called a double-blind, one is called a single-blind. Both require more than one person. You find a situation where a lot of dog handlers are out there on their own with no help. So they place things out for the dog to find themselves. They already know where it is. That's not blinded at all.
Unfortunately, those dogs seem to have a greater problem than dogs that are trained and tested by placing things out by someone else, where the handler doesn't know where it is.
CONAN: Oh, because the handler then sort of tips off to the dog if he set out - hid the stuff, he knows where it is.
MYERS: That's right, and in many cases it appears not to be at all intentional. Dogs pay a lot of attention to our behavior. If you look at something for just a little bit longer than something else, the dog almost always will go over and investigate that.
CONAN: I see. Lewis Katz is the John C. Hutchins professor of law at Case Western Reserve School of Law, where he writes about the legal issues surrounding the use of detector dogs. He joins us now by phone from his office in Cleveland. Nice to have you with us.
LEWIS KATZ: Thank you for having me.
CONAN: And there is some history in Supreme Court decisions that the alert of a detector dog constitutes probable cause.
KATZ: Yes, in 1981, the Supreme Court said that a dog's sniff is not a search. It wasn't essentially to the decision in that case, but it has been the basis for thousands of decisions since 1981 upholding police use of dog sniffs in various contexts without prior judicial approval or without any scrutiny as to why they're targeting the individual.
CONAN: And have there been restrictions placed on that subsequently.
KATZ: Certainly not at the Supreme Court level, which is why these two cases pending this year are extremely interesting and crucial.
CONAN: One involves a car and one involves a private house, and I gather those are viewed rather differently.
KATZ: Well, I hope so, but that's anticipating how the court may act. Ordinarily the court says that a home is entitled to the greatest Fourth Amendment protection, and the court has said in the last 10 years that the use of any technology on a home, such as in that particular case a thermal imager capturing heat waste from the home, constituted a search, which required a warrant issued by probable cause.
CONAN: They were looking for the heat signature of somebody in a grow house. You're obviously using a lot of heat lamps to grow the marijuana.
KATZ: Yeah, but incidentally and often lost is the fact that they were using it on the whole neighborhood.
CONAN: So it was, in other words, look at not just a place where they might have some suspicion - not apparently enough to get a warrant in advance - but at the whole neighborhood so if somebody else's house had come up, they would've been busted too.
KATZ: Presumably they would have done so, yes.
CONAN: And as you look at the situation of a car, why is that considered different?
KATZ: Well, the court has said that there is a lesser expectation of privacy in an automobile, and therefore the court has said that any validly stopped automobile or a parked car may be subject to a canine sniff without any oversight as to why they're targeting an individual car.
Obviously they can't bring the dogs around every time they stop a car. So the question arises, why should they target this particular individual. But under existing law, the police don't need to present that justification.
CONAN: Larry Myers, let me ask you, are experts on dogs and their ability to smell, olfactory senses, are people like you ever asked to testify before judges to say how good or bad a dog might be?
MYERS: Yes, frequently, actually, not just me but there are a few others, a few scientists who study this sort of thing and certainly a number of current and retired handlers and trainers are found in courts on these cases.
CONAN: What kind of questions do you get asked?
MYERS: Well, usually questions that can't be answered definitively one way or the other.
(LAUGHTER)
MYERS: How good is this dog? How good is this handler? And the best that you can actually do is look at the records and any video recordings that might be available, that sort of thing, and make a general assessment. But without actually testing that team objectively, which is rarely allowed, I can't make a definitive statement, I can simply state that this individual has met the more or less reigning standard, or this individual has not. And it's based largely on the records of that team.
CONAN: And Lewis Katz, let me ask you: What constitutes probable in the probable cause thing? If a dog is right, say, 50 percent of the time, is that probable cause?
KATZ: I should think not. A dog that is correct 50 percent of the time means that there are 50 percent false alerts. And I raise the question whether that is adequate probable cause. But I think equally important is the question before that, which the police don't need to answer now. Why did you alert - why did you use a drug dog in this particular situation?
You know, it doesn't only apply to cars and houses. Drug dogs are used in some schools to sniff around children, see if the child is carrying an illegal drug or a prescription drug.
CONAN: Or their lockers, presumably.
KATZ: Well certainly lockers and automobiles. I think it's really significant, and many school districts don't do it; to use a dog on a child, considering the child may be afraid of dogs or any other number of situations.
CONAN: Let's get a caller in on this. Randy's with us calling from San Francisco.
RANDY: Yes, hi Neal.
CONAN: Hi.
RANDY: I'm an attorney. I do litigate these issues in terms of dogs and probable cause. I was wondering if your panel wanted to address - there was a study done at UC Davis that involved 16 teams. Half of them were for narcotics detections, half of them were for explosive detection. They used a brand-new church to avoid the residual odor problem.
The study was explained to the participants that they were going to use noise and other distracting devices to see how the dogs functioned. They had eight rooms. Each of the 16 teams was to go through the eight rooms. Two gentlemen showed up with attache cases handcuffed to their wrist. One had a pound of C4 explosives. The other had a pound of cocaine. All 16 teams went through eight rooms. All 16 teams identified at least one location in which the dogs alerted.
When the study was over, they asked the person running the study, doctor, how did we do? And she told them that they had not planted any cocaine or any explosives in any of the rooms. What the study was really testing was the fact that the handlers - and these were top handlers from all over the country - were involuntarily cueing their dogs based upon their expectance to find the contraband due to the setup of the experiment.
CONAN: And...
RANDY: And this raises a specter that without proper training, dogs are involuntarily responding to their masters and really rubber-stamping their masters' intuition, which is a traffic officer making a stop and then becoming suspicious.
CONAN: Larry Myers, are you familiar with that case, that test, and is it definitive?
MYERS: That was - I think - I believe all of those were actually narcotic detection dogs. If it's what I'm thinking of, that was by Lisa Lit in April 2011. And there were 18 teams. Seventeen of them ended up with more than an aggregate of 200 false alerts between them. One team, in fact, did not false-alert at all, suggesting, one, certain training and certain handling can work, but it also suggests that dogs are exceptionally well attuned to the behavior of their handler.
It's interesting that that happened 100 years after a similar thing had been seen with a horse called Clever Hans. That was first published in 1911, so we've known about that principle for a long time.
CONAN: Yeah. Clever Hans could solve mathematical problems that were unconsciously tipped to him by his handlers.
MYERS: Right, right. And dogs are the same.
CONAN: Dogs are the same. And, David Latimer, how do you correct for that in training?
LATIMER: You correct for that by doing double-blind training and testing.
CONAN: And maintenance as well.
LATIMER: And maintenance as well. And the important thing about that study is, as Larry mentioned, there was one dog that performed no false alerts. There was another dog team that performed very few false alerts. So just don't throw the baby out with the bathwater, in other words. If we handle this properly, dogs are very important and useful to the law enforcement.
CONAN: But, Lewis Katz, clearly this has to be treated with some skepticism.
KATZ: I agree. I'm not sure the science is as precise as we might like. And the dog - in many courts, the handler is - handler's word that the dog alerted is simply accepted by judges as the basis for probable cause, rather than inquiring as to different kinds of alerts. And, you know, sometimes a dog wagging its tail, in some of the cases, has been indicated as a positive alert. So I think we have problems there, as well. It's not just the science. It's how the courts react to what they're told.
CONAN: Randy, thanks very much for the call. Appreciate it.
RANDY: All right.
LATIMER: Can I add something here?
CONAN: Quickly if you can.
LATIMER: It's not the science, however. It's the courts, more of the problem, than it is the science. Science is solid behind dogs being able to detect odors.
CONAN: It's the courts being...
LATIMER: Courts are not being - are not educated enough to ask the right questions and perform the right kind of testing in a - such as a Daubert kind of challenge.
CONAN: We're talking about detector dogs. You're listening to TALK OF THE NATION from NPR News. You just heard David Latimer, police chief and dog trainer in Harpersville, Alabama. Lawrence Myers is also with us, associate professor of animal behavior and sensory physiology and medicine at Auburn University. And Lewis Katz, John C. Hutchins professor of law at Case Western Reserve School of Law in Cleveland.
And let's see if we can get another caller in on the conversation. Let's go to Greg(ph), Greg with us from North Park in Colorado.
GREG: Yes. Hello.
CONAN: Hi.
GREG: Well, I just wanted to - I'm a little bit off-subject here. I'm not - I don't train law enforcement detector dogs, but I train dogs for conservation work. And I just wanted to kind of throw that out there as something with very similar methods to what the law enforcement use, but we use them for things like in wildlife research.
For instance, I just finished a cougar scat study in Northeast Oregon where we - the dogs found the scat. We collected the scat, got DNA from the scat, and we're able to do population estimates from that. But they've been used from everything from desert tortoises to different types of scat - whale scat, believe it or not, and quite a few nesting birds, ground-nesting birds. There's been quite a few applications in conservation work with dogs.
CONAN: Excuse me. Whale scat?
GREG: Whale scat, yeah, believe it or not. So whale scat floats, and they had the dogs on the front of a boat, and the dog would basically point with his nose to which way the smell was coming from, and the boat driver would point the boat that way, and they would eventually find where the whale scat was, yeah.
CONAN: You learn something new every day.
GREG: Every day. Yup. It's pretty phenomenal what they can do.
CONAN: Thanks very much for the call, Greg. Appreciate it.
GREG: You bet. You bet.
CONAN: And we just have a minute or so left, but, Lewis Katz, as you look forward to this decision by the Supreme Court in these two cases, do you expect that we'll see some limitations - or could we see some limitations - on that estimate that an alert by a detector dog constitutes probable cause?
KATZ: Well, I think the issue may be that the court says walking a drug dog around a home constitutes as search, and therefore the police have to get prior approval from a judge in the form of a warrant. And I think that would be a major step, because under the laws that exist now, and if they don't rule that way, police can - and they may not do it often, but they could - walk a drug dog through apartment corridors just checking on every apartment on the floor. And I think the room for abuse is too great here and are adherence to Fourth Amendment values in protecting our homes is sufficiently important that I think the court may draw the line there.
CONAN: Well, Lewis Katz thanks very much for your time today.
KATZ: Thank you.
LATIMER: I agree with Lewis on that point, by the way. Our Fourth Amendment are important. However, one thing I disagree with him, I have never searched or have I ever seen anybody search, a child with a dog. That's a new one on me.
CONAN: And that's David Latimer. Thank you very much for time today.
LATIMER: Sure.
CONAN: And Larry Myers. We thank you for speaking with us as well.
MYERS: Appreciate being here. Thank you.
CONAN: And we'll await the Supreme Court decision coming up some time in the next few months. After a short break, though, we'll launch this year's series of conversations with the people behind the Oscar-nominated feature documentaries. First up, "Searching for Sugar Man." Stay with us. I'm Neal Conan. It's the TALK OF THE NATION from NPR News. Transcript provided by NPR, Copyright NPR.