에피소드 10: Simon Gadbois

and I strongly believe that uh to really understand behavior and that's my bias I have to admit um all these perspectives are important uh uh because they're not competing with each other they they actually complete each other and I find that a little bit uh unfortunate that I think a lot of people in the dog world see uh ethology as the opposite of psychology or you know the the the good old nature nurture debate kind of thing which you know in science now we don't even think is a debate at all actually it's but it was a quite fierce debate at some point back in time correct? yeah before I would say before the 70s that was you know the the big war between ethology and psychology was about nature and nurture but then both groups both groups kind of realized they were both wrong and both right about some stuff right and it's really Robert Hynde and and and and honestly I have to say here uh for full disclosure uh he's part of my academic uh genealogy it's Robert Hynde from um Cambridge that kind of said well wait a minute you know psychologists you have amazing points and ethologists you have amazing points and look these things go together perfectly well if you acknowledge that behavior is not just nature or just nurture and it's a complex combination of both um and it's a little bit worrisome for me to see the the new wave of behaviorism especially in the US but it it it's pretty much you know the the new wave of behaviorism you have a lot of different types of threads not in academia because it's not really a thing since probably the the 80s or 90s mm-hmm but mostly in the dog training or animal training world um and I see more and more of it uh you know like theories that were very popular 10 years ago for instance that now are being criticized um you know from I would say a behaviorist perspective um you know from I would say a behaviorist perspective well for instance the the work by Panksep or some of the people more from the neuroscience perspective because it's biological now it's it's been vilified and it feels like we're going back in time like you know like some of the stuff that was going on in the 60s and 70s and that's very very strange to me and I don't know where that appeal of the radical behaviorism comes from I have nothing against behaviorism if it's what we call operational behaviorism which means that it's a behaviorism that's open to talking about emotions for instance on the brain and correct uh you know cognitive processes but the radical behaviorism that that really uh you know wants to avoid any of those concepts

is is is is more popular than I think it should be right now I just don't get it yeah I I have to agree with you as I mean I I travel even with call it I travel quite a bit somehow I find ways and I teach and I have a school and um it's what what you're saying is spot on like the for whatever reason somehow that type of radical behaviorism has stuck very very strongly in in the dog training world and and it almost doesn't even matter well we know there is so many different ideologies and camps if you want to call them this way but it really doesn't even matter uh one will pick up one side and another extreme and so on but ultimately they'll somehow go back to to radical behaviorism and and after again just as you said so many years we we we know better right now I mean we have evidence and we we and it it's so much more interesting actually well yeah and it you know I I think one of the reasons is that if you really focus just on conditioning to explain well actually that's the thing it I think it's the the pragmatic or the practical appeal of conditioning that that makes people say okay that's all I need right and it's true if you're a dog trainer uh in a lot of the work we do in my lab here the the main tools in the toolbox are obviously from basic conditioning principles to train our animals to do what we want them to do the problem becomes when you claim that that's all you need correct to explain animal behavior and and that's where the academic world has gone way beyond that kind of thinking um and for me since we were talking about my background earlier I spent some years studying wolves for instance and uh but before that I had worked with rats in some kind of neuropharmacology kind of stuff I had worked with pigeons and in spatial learning and I was kind of a behaviorist maybe not radical but at least operational or you know more more relaxed type of behaviors and I remember the very first day that I spent watching uh observing some of the wolves I was going to work with for almost 10 years at that point I just said at one point wow this is all limbic and what I meant by this is that when I was watching them interacting with each other there was not much that conditioning and a pure radical behaviorist perspective could have explained um it was raw emotions the limbic system was almost behind everything that I was seeing there was not much even cortical in the in the in terms of let's say social cognition for instance um and I'm not saying that they're brainless animals on the contrary I think they're really smart animals no doubt about it

they have very uh great uh both instrumental and social intelligence but you know if you reduce yourself to basic conditioning principles to explain animal behavior of any species you're missing a lot a lot right and you and there's so so many interesting topics hopefully we we can touch everywhere because I really would love to um you you're you're based in Canada and and that's kind of probably where the are you born in Canada or uh yeah so that's probably I mean it's easy to well not easy but you have an easier access to to study wolves and and to uh uh probably yeah it's uh it's yeah it's true in a sense yeah because it's part of the culture in in a sense too I mean they are in most well well well actually I was going to say in most provinces not not really but they are in a no actually they are in most provinces uh interestingly I'm actually in uh in uh in a province where there there are no wolves uh in fact the maritimes New the Maritimes, New Brunswick, Nova Scotia, and Prince of Rhode Island, there are no wolves. In fact, they're not even in Newfoundland, I believe. But otherwise, you're right. You know, the wolf is an animal that is part of our, yeah, is part of our culture. It used to be the case in the U.S. as well, especially the western part of the U.S. actually at one point almost everywhere on the continent, but certainly it is still part of the culture of some of the northern states like Minnesota, Michigan. But yes, it's a little bit different here for the access. As I say this, there's probably more wolf researchers in the U.S. than Canada overall. It's probably a question of obviously our right population. We're a tenth of your population in the U.S. So that probably explains part of it. But I don't know. In fact, a lot of the wolf researchers that are Americans come here to study wolves, right? David Meech studied wolves on El Smeer Island here in Canada. I really tried to have him at some point on the podcast and he agreed. But then I guess just the agency he's working for still, it's not, I guess he's not allowed to yet. But yeah, there was something to do with his job that he was not allowed to. Do you ever think of going back to wolves? I mean, I know it's a completely different lifestyle and it's just not a... Yeah, no, actually. I'll tell you two reasons. You're right. The practicalities of it, especially since I have kids, you know, this is like months and months away. It's very demanding in terms of the fieldwork, the time. And yeah, no, not really.

The other factor I have to say is that when I stopped, I was actually relieved because as much as I like wolves, they're not necessarily my most favorite canids. I worked also with foxes and coyotes and I absolutely love coyotes, actually. I think coyotes are fascinating. But the problem with wolves was kind of some kind of aversive conditioning in the sense that the humans that work with wolves are really difficult to deal with. There's a lot of big egos. And I think it's in part, there's two things there. I think there is the fact that it's a limited resource. You know, it's like you pointed out, it's hard to get funding. It's hard to get access to them. There's a lot of competition. So there's that aspect of it. And the other one is, I think it's what I call a high profile species. I was at a conference as a graduate student in 1996. And we ended up being a group of people that we were all complaining as students about the politics of research. And we noticed that we had one thing in common, all of us at that table, is that we were all studying high profile species. So primates, cetaceans like whales or dolphins, wolves, right? And I think there's something to that, unfortunately. Those species are kind of magnets for, you know, people that are there for the wrong reasons, basically. More for the attention of the species they're studying and not really for the passion of it. And I remember when I ended the wolf research, I remember thinking, I'll just go study chipmunks. Because I think chipmunk people are nice to each other, right? I mean, nobody competes for chipmunks. Everybody's happy to work with chipmunks. Why does that remind me of dog training communities? Well, there's that too. So then, where are you right now in terms of what is super interesting? What are you into right now? Yeah. So in 2007, we had the Wolf Research Center here attached to the Lausie University that facilitated a lot of the wolf research. And there were some great people that worked here, like Fred Errington, John Fentress, Peter McLeod. But the center closed. A lot of those people retired. So I had to reinvent myself a little bit. So I worked a little bit on coyotes for a while. But then I started actually developing a research program with wildlife conservation canines. And basically, this is kind of what we've been doing in the last, let's say, 12 years, almost 15 years at some level. And it's to study applied areas of dog olfaction. The two main ones that we've done over the years are biomedical applications, mostly for detection of hypoglycemia in children.

More recently, with Laura Kirroja, my graduate student, currently anxiety and stress detection in people with PTSD. And again, the wildlife conservation canine stuff, where we've done all kinds of projects, actually, training the dogs to look for ribbon snakes that are endangered here in Nova Scotia, wood turtles, bat and bird fatalities at windmills. We're getting into new projects now on detecting nest and eggs of some ground nesters, birds. Also looking for owl pellets to look at a project to see how much owls get contaminated by lead. Wow. Yeah. Oh, wow. So what kind of dogs do you, how is the selection and how do you get there? Yeah. So the way it works here is that we work with volunteers from the community. So we don't have a kennel. It's people that volunteer to dogs to work with us. And some of the projects that are within those areas that I mentioned are methodological things. Like we've looked at different things over the years, like different training methods, for instance. Now we're getting into dogs that are sign trackers over goal trackers, for instance, in training and how that may be relevant to some of the skills, especially for biomedical alert. But we basically get whatever people volunteer. So for a while, I would say 90 to 95% of the dogs we were getting were border collies. Interesting. Because border collie owners are desperate to find a job for their dogs. Yeah. Yeah. Yeah. And we we went with it because. I mean, border collies are weird there. It's a bit all or none with them. They're either a bit too ADHD. They don't really focus or they require a lot of work or there's some kind of neurotic behavior or being in the way some kind of OCD like behavior. By the way, I don't typically like using these labels for dogs, but I think. No, no, I totally understand what. Yeah, absolutely. But when you get a border collie to first of all understand that they have a nose because a lot of them learn to use their nose here. Even the owners will say, I don't know what you did to my dog, but he sniffs everything now when they get the idea that they have a nose. What's remarkable with them is their work ethics. Again, another term that's very vague, but I think we all understand what what it means. The only issue sometimes is they may some of them may get bored with time, the repetitiveness of what we do in the lab. But again, a lot of them will just go like, yeah, no, I mean, this is my job. This is what I do, and I will do it perfectly for as long as you want me to do it. So for many, many years we were mostly a border collie lab.

It's only in the last five years with my new graduate student that we kind of opened up to other breeds with some interesting successes. So, yeah, we're a little bit less focused on border collies, but I would say it's still about 70, 75% of the dogs we get here. But it's still probably interesting to work with other breeds just to experience the difference and what takes you, right? Yes. And, you know, I had a border collie at the time when we started. So I was a little bit biased towards that breed anyway. But in the meantime, I got a red golden retriever. So this is not your regular, you know, teddy bear, fluffy couch potato golden retriever. These are very slim, very slender, very active working breed. In fact, the one I have is much more intense than any border collie I've ever had. So, and they have a magnificent nose. This is why I realized when I got IV six years ago, yeah, okay, there are some noses that are better than others, no doubt about it. And they're definitely also one of the breeds that it's so highly in tune into working with the human that it's... They love to please, right? They'll do anything to please you. So it's almost annoying sometimes how much they love to please. I remember working in the early nineties at guide dogs for the blind in California. And at the time, I don't know what they have now, but at the time we did have labs, goldens and some shepherds. We were starting to get away from shepherds just because they were excellent to work with. But the people that were coming to get dogs, they were just not physically capable to go with that pace. So all the, like some of my favorite dogs that I worked with were goldens. And some of the most ridiculous memories that I have are still with goldens. Because I would work a dog and granted it's kind of in the middle of the training and it's serious work, of course. We're right about to approach the curb. The dog has to pay attention. And next thing you know, there is somebody just around the person that walks by. They have to greet them. And with a full blast wiggle and bow and like just, just check out like an instant to where it makes you laugh so much. So yeah, it's very, I mean, they're cool dogs. Tell me, I'm interested in how, what do you, how do you, what do they get paid? Like when, with the detection work, how, what is the reward normally? How? So in the, in the lab conditions that, that, that we have, because a lot of the work we do here is, is not actually, including for the biomedical stuff, is not really, it's not with the people being around.

So it's with samples of breath, for instance, we just want to demonstrate that from breath, they can detect just by scent, different states, either in hypoglycemia or stress. So they, they basically sniff vials that are in special containers. And we, we, we, we, we used a clicker. I don't know why some people started saying that I was against a clicker. I was never against a clicker. I, I think it's a great, great way to train dogs. And we, as a primary reinforcer, we use special treats. It's actually, it's a mixed bag, literally of different value of treats, you know, from regular kibbles to, to more exciting ones. Just aligning here with the Rescola and Wagner model, you know, that the, how surprising the unconditioned stimulus is, is actually a good idea. So, yeah, the reinforcer basically is, we have a secondary and primary reinforcer, and that's how we've been training. As I say that, we have, we assess all of our dogs initially to see what they respond to the best. So, over the years, we've had a few dogs that don't tolerate the clicker for some reason. So, they, we, we used only primary reinforcers. We have a few dogs also that respond very well to play, better actually than food, as a reinforcer. And with those dogs, play was used. But as you know, play as a reinforcer can be a little bit problematic if you don't have a very good discipline, you know, in making the play about very short. Especially some of those breeds like those red, golden retrievers that will, you know, they get into play and then you can't snap, snap, snap them out of it that easily. But we, we have the policy in my lab to adjust to the dog, not force the dog to adjust to our ways of doing things. Which is so typical for us, for majority of trainers. It's like we, we learn a way or we get excited about a way and, and we almost forget that. To pay attention to the dog and see, maybe that's not the best way for that one dog. Yeah. And it's, it's also a little bit difficult, honestly. You know, when we started working on this kind of stuff 15 years ago, publishing some of our research was difficult because we would, as you're supposed to do, we would acknowledge that, you know, this dog was trained this way and this dog trained this way and this dog trained this way. If you send this to psychology journals, you know, that read mostly about pigeons and rats being trained always exactly the same way, usually by computers on top of it. It doesn't go well back then, at least there was a lot of pushback, like, no, no, no, you have to train the dogs all the same way.

Otherwise you're comparing oranges and apples. And I would say, well, not necessarily. They were trained to a criterion. That's all what matters. How we got there is in the applied world of training, at least kind of a detail. Correct. Which, which kind of goes to, to what you're teaching right now, learning and motivation, you know, like if that's the motivation that really, if that's what motivates you, it will be, it will be almost a shame to, to push it away and try to motivate with other means. It's very interesting. You know, we, I, I actually, what I fight against often are students that are, you know, are coming from the, you know, very, very formal research methods kind of perspective. I used to teach that course here, actually, so I understand exactly how they think about these things, but I'm the one that have to remind them. Yes, but this dog is not responding to this right now really well, and we're losing him. So what are we going to do? Are we going to continue to try to shove down their throat this kind of method that it clearly don't handle well? Or are we going to try to see what gets that dog to tick right to react properly to what? And usually the common sense gets them to say, sure, we'll adapt to the dog. But it's funny how the go to thing is, well, this is how we do things, you know, and then we, they want to apply that method to all dogs all the time. And it would be beautiful if it was working, but it often doesn't. Yeah. In my opinion, motivation trumps everything. And like when, when I teach and where I go, I also have a, I have a school for dog trainers. I just started like about two years ago. It's an online school and then they come for certification and we do some hands on and do this kind of Zoom calls. And one of the big things that I really, really insist and keep repeating is that motivation is priority. Finding motivation, maintaining motivation, that it's a, it's a healthy motivation because you can have a, you know, you can have some dogs can go way crazy to where they cannot really function. And then it becomes counterproductive, but getting like, in my opinion, most dog training, even, even today revolves around how we're going to accomplish a behavior, how we're going to mark certain things and when we're going to reward. And when we're going to reward, but not taking under consideration that all this mechanical work, dogs really look at us like, hey, are you alive? Do you like what you're doing? Are we going to actually do something together? Or are you really treating me like a little machine that I'm going to sit down and come?

Or are you? And why do you want to do it? Is it because you're afraid to do it? Afraid not to do it? Or is it because... There's so many nuances of motivation, correct? But it really comes down to... Because if there is motivation and if the trainer understands how crucial it is, you may not be the best trainer. But you have a dog that's willing to go on and figure it out for your lack of handling skills just because you've tapped the right motivation of the dog, correct? Absolutely. And two things here I won't comment on. One is I think we have to get off this idea that there's no breed differences. I know this is very trendy right now and some people are pushing this idea, but it makes no sense even from a biological perspective. If you understand how artificial selection works, if we've managed to create different phenotypes based on morphology, so size, et cetera, et cetera, behavior is a phenotype too. So obviously there are trends, at least within breeds, that differentiate border collies, let's say, from St. Bernard's. There's no doubt about it. And it drives me nuts that some people refuse to see this based on whatever kind of principle, in part actually motivated, I think, by breed-specific legislations. But that's another problem. I think that acknowledging as breed differences in behavior and behavioral profile doesn't necessarily have to go with that issue. That's a completely different one. So, you know, when I was working with border collies, I remember some people were talking to me about their dogs and we would have conversations about how to train them. And they would say things like, I don't know, well, you've never trained a malinois, have you? And then I would see the trap coming and thinking, oh boy, okay, yeah, you're right. I haven't. So, yeah, I know this is a different reality. So then we would talk about, you know, the quirks of malinois, which I know you know quite well. And, you know, over the years I discovered that, yeah, indeed, obviously there's always individual differences, even within breeds, and I've seen it quite often actually with border collies. But there are some huge differences. Now that I'm also more open to hunting breeds, which I wasn't initially, I just discovered a completely different world, especially for the wildlife conservation canine stuff that we do, where I'm starting to see now, you know, we've created all of those hunting breeds for very, very specific purposes sometimes. Half of the work is done there, just in terms of the selective breeding that was done. And I never really took advantage of it. And now I see that I should have.

You know, there are some hunting breeds that we've used that were complete disasters, but when you look at what they were bred to do, then you realize, yeah, what was I thinking? So, you know, that's one part, the breed part. The other one is what you say about motivation resonates with me very well, because I think that, again, this is an effect of behaviorism. Radical behaviorism, they never wanted to talk about what we call hypothetical constructs, right, like cognition, like emotions, and motivation was one of them. The problem with this is that this has excluded all of the new modern theories on motivation and learning that are all about what we call incentive learning or incentive motivation. And it's a beautiful world because in all of those theories, and there's so many people involved in this, Berridge, Panksepp, Ballin, Dickinson, Schultz, etc., etc. What they're basically telling us is that we can motivate, we can, sorry, we can learn or teach or train motivation. We're missing a huge tool in the toolbox, which is this idea that motivation is something that we can actually transform. As trainers, we have the ability to actually transform and direct motivation. It's not just about the intrinsic motivations of some individuals. It's also that you can teach them to become motivated by some things. In a way, we already do it with the clicker. I mean, it's a great example of how you're basically telling them that the clicker is announcing the potential, at least, of a reinforcement. And that's a lot of the magic of the clicker. It taps right into the motivational system of the dogs. This has been really well demonstrated with rats, with humans. Even one study by Burns, I believe, on neuroimaging is showing that indeed some of the parts of the brain that are known to facilitate learning get activated by conditioned stimuli, such as the clicker, for instance, or even praise, for that matter, works as well. So, just to interrupt, we'll go back to this. But since we are right on it, I want to ask your position on this, because I go back and forth. I mean, I do have an opinion. Condition reinforcer. There is a notion that I tend to subscribe to more. And that is that it's not necessarily a condition reinforcer, but it's like a signpost. It tells you what happens, but it's a signal more than actual having its own properties. Do you understand what I'm saying? I think what you're talking about is the difference between, let's say, the clicker as a secondary reinforcer or conditioned reinforcer and the clicker as a marker. Is that possible that that's what you're...

No, no, just the fact that as a conditional reinforcer, does... Does... What are the... Basically, what are the reinforcing properties of the clicker? Or is it really just the primary reinforcer and the clicker's signals? You know, just like a sign that says there's going to come a right turn, but nothing further than like that sign that says right turn doesn't necessarily have any reinforcing properties. So that's a huge area of what I was talking about, these incentive theories of learning. And, you know, there's been some debates about this. I've been dragged in some of those debates with dog trainers in the past. What's really interesting about this is that I think part of the answer is it depends. It depends how you use the clicker. It depends what your goals are. And I think this is where a lot of that discussion, and it took me too many years to understand this, are going in the wrong directions a lot of the time, and we talk past each other. One of the ideas, for instance, is that the clicker predicts that a primary reinforcer is coming. The key word is predict. And the question is often, does it need to predict at 100%? This is the whole idea about a one-to-one ratio between, you know, the clicking and then delivering the kibble. And it's really interesting because if you look at that literature, it's not black and white at all. Not at all like a lot of people are actually saying it is. It is not. It doesn't matter if it's rats or other species. It's not at all black and white. And there are a lot of theories that are relevant to this idea. And again, it goes back to it depends what you're trying to do. Like ultimately for a dog trainer, it really, bottom line, it does make a difference. And now looking at from what kind of glasses we put on, we can probably start looking at it as, well, it is a conditioning force or not. It's a signal or not. You know, but the bottom line for a dog trainer, you do need. It definitely dogs are looking for dogs. You know, so, oh, thank you. Yeah. Yeah. Yeah. Well, so, you know, part of that debate also has to do with if you are okay with the concept, in part, at least, of intermittent reinforcement, which is another thing that now is vilified a lot in animal training for reasons I don't understand. And it's really interesting because a lot of the people that are against are actually Skinnerians, which is fascinating because much of Skinner's PhD thesis and the content of his book, The Behavioral Organisms, is actually on intermittent reinforcement.

So the fact that they turn against it fascinates me, especially when you look at what intermittent reinforcement is supposed to do, what the point of it is, which is to often increase the responding, but also a huge resistance to extinction, which, by the way, is often a huge problem with medical alert dogs that stop working or stop responding after a while. And the main solution, if not the only solution to prevent this with these dogs is to make sure that they are on intermittent reinforcement. Because a lot of the stopping of responding that occurs in these dogs is because people don't keep the contingency, and it's also a very rare event. So two things going on. It doesn't happen very often that you'll have a hypoglycemic crisis or that you will have a panic attack. So the behavior has a tendency to extinguish, number one. Number two, again, when people are going through the event and the dog responds properly, they often are too busy dealing with their blood sugar levels or their panic attack that they don't actually reinforce the dog. So obviously, after a while, the dog stops responding. It's like, well, okay, I don't see the point of this. And they stop doing it. This is a real, very well-known problem with people that have those dogs or training those dogs, although nobody really is talking about it openly, is that you need a lot of maintenance training to keep those dogs doing their job reliably. After a while, you're guessing. After a while, you're guessing. You're hoping they're doing it. You're hoping they're paying attention. But it's not guaranteed. And it's interesting because you go back to the rationale of intermittent reinforcement, and boom, then you realize that the whole point of it was precisely to deal with these kinds of situations. And yet, a large number of dog trainers are against it because they will say, for instance, that it frustrates the dog. Thing is, if you are implementing intermittent reinforcement and you see frustration in a dog, you're doing it the wrong way. True. That's what we teach to our students that shape pigeons and do that work with rats in Skinner boxes. You're going too fast. That's well-known. Even Skinner was clear about this. If you change the schedule and you increase the time between the good behavior and the reinforcement or the ratio, you have to do it very, very progressively. And what most people do, including my students, they go way too fast. They go from 100% to 50%. And then they wonder why the dog is frustrated. But that's because any dog would go like, what the heck is going on here?

This, you know, you used to give me a kibble every time. So. That would be Las Vegas casino. Everybody breaking the machines. Agree. You know what I think with this is that, like, from my experience, there is, I do see how some trainers in some situations may insist on continuous schedule. And that would be very different. And what we are demanding. And so, for example, like, if, and I do a lot of high-level competitions. So there is behaviors that are what I call high maintenance and a little bit almost requiring something above the natural response. And in those, I find that because it's so demanding, that continuous schedule is, it has some special value because it kind of, I made it. You know, that question mark. You know, I'm not sure if I'm a player, but I'm a player. I don't quit. So they both have the interesting values. I completely agree. And it goes back to this idea. It depends what you're trying to do, right? And the theory behind intermittent reinforcement, let's say, from a neuroscience perspective, is that now we know from the neuroscience of learning that it's not really the reinforcer that's the key as it is consumed. It's the expectation or the anticipation of the reinforcer that gets the learning to work. So intermittent reinforcement addresses this in part if it's relevant to what you're trying to do, where you get the animal a little bit on its tippy toes, right? And you can see it actually with rats when you put them on some schedules where the rate of responding will increase. If that's what you're trying to do, intermittent reinforcement works really well, especially if it's fixed ratios, I believe. Yeah, it's where you get the rat going like this. And your analogy, maybe accidental, about Las Vegas and gambling is totally right, right? Because the people that program those gambling machines, especially the slot machines, know exactly those rules. Exactly. They don't want you to get angry. They want you to keep plugging. And at the same time, they know you're not going to have the payoff every time. They want you to have a high response rate. They know exactly when to deliver the reinforcer. And it gets literally people addicted. Now, I'm not a proponent of the idea that dog training is about getting a dog addicted to anything. I don't think that's the point. But to go back to the link between primary and secondary reinforcer, I think part of the idea there also, and I got myself apparently in trouble with this with a known dog trainer a number of years ago at a conference.

I had pointed out that actually, theoretically, when you do clicker training, you're basically doing pure classical conditioning initially, where you just pair the click with the food, click, food, click, food. The dog acquires this. And there is a literature on this. In fact, it's in most of the textbooks that actually, in a sense, the clicker by itself has acquired a reinforcing value. And removing, not all the time, but removing occasionally, even if it's accidentally, the delivery of the primary should not affect learning. If you remove it completely, eventually, that will break. But this idea that you must always click and always follow with the primary, there's actually no strong data for that. If you stop completely, just use the clicker by itself, eventually, maybe. But I say maybe only because, you know, a lot of what we do, especially in the kind of agility and more complex kind of competition stuff that you do, some of the stuff I do in the field as well. You know, dogs can't care less if I'm clicking or not. They are so self-reinforced by the work that they do. You know, sometimes I have my beautiful, fantastic liver treats in the field. And I want to, you know, we just found a turtle. I'm all excited. And they look at me, it's like, I don't have the time for this. I need to go back and find another turtle. I love it. I love it. That's the ultimate training. Like, really, ultimately, what we want to accomplish is exactly what you just said. It's like, you, me and my dog, we're going to go and do the thing that you and I love doing. And all these blocks, they were just blocks. Yes, we had to be precise as we had to do. So, but there is a point in time where, like, again, like I go so much and people think, oh my God, he must be so spot on with every single time that something happens. There is a reaction and there is always acknowledgement. And they get so surprised. It's like, oh, you didn't mark. You didn't release. You didn't. It's like, guys, I have worked with this dog for five years. Like, I can, I mean, the clever Hans story. I mean, we don't need nothing anymore. We are just looking at each other and I can tell him to come to me probably in 50 different ways. It's, there is a point where that basic language should flourish and then we just get to enjoy what we like doing. If assuming, again, that motivation, we have tapped into the right motivation where I kind of wanted to go back there.

One of the interesting things for me when I train, and it doesn't matter if it's detection, because I also go and do a lot of detection for military and around the world or search and rescue, whatever it is. But there will be dogs that will come that will be from different breeds selected genetically to do a specific job. And if we, as trainers, put them all in that same loop and say, okay, well, that's the reward, that's the reinforcer, that's the, instead of trying to find what is genetically reinforcing to that breed. Because if we, if we can figure it out, that genetic program, this is what I'm made to do. I don't know why I'm getting so hyped up about it, but that's what it takes me. And next thing you know, if we find that, that dog is like, I really like this guy. I don't know why, but he is tapping into something that nobody else touches in me. Right? And when we start to look into, okay, that, let's say, okay, it's a border collie. I mean, yes, we can give it the toy. Yes, we can give it the food. But ultimately, the whole reward can go a little bit into some stalking, some backing up, some forward, some, okay, grab the ball now. To where all those instincts, all that selection gets into play. And when that happens, that's the most brilliant stuff. Like if, and I'm sure you know, like Chaser and John, and unfortunately I never got to meet him, but I talked to quite a few people that knew him at the time. And every time I watched some of the videos, ultimately, the way that dog responded was not necessarily, well, he liked to play. He liked to play, but also he tapped into the typical border collie genetic makeup. It's like, no, that's how we're going to play. You're going to stalk it. You're going to chase. You're going to back up. You're going to go to, you know, then you have a different type of dog that maybe he needs to search a little bit. Or maybe it's one of the bully breeds that they just enjoy to hold on to something for a long time. So if we start to look into what you like and apply it in the training, it changes everything. But I want to go back a little bit. You mentioned the goal and sign tracking and something with the detection. And I know the, like, just give me a little bit more there because you said you're doing a little bit more into sign tracking. I don't know if I missed it a little bit. Yeah. So first of all, a lot of what you talked about there, I think for me is this idea that, you know, you can teach motivation. I think the idea is to develop intrinsic motivation in dogs. That's the, you know, very important thing.

And I think that's what we do as trainers in a way is we get them to like what they're doing. And sometimes they will do it for free if they really get into it. And that's the beauty of it. Like you said, it's, and a lot of the time you're right about that too. It's to bring to them more natural behaviors. You know, one thing we've noticed in my lab here is that we do, we do medical detection with, you know, kind of a sterile lab, you know, a floor that's washed every day with, you know, like glass vials and stainless steel containers and that kind of stuff. And it's really boring for dogs. And a lot of the time when we see their performance with time go down, especially dogs that have had experience in the field. And then what we'll often do is we'll start burying those samples into a container that has wood, wood pellets for wood stoves. And then suddenly the behavior comes back. And it's kind of intuitive because, and the performance goes back too. And it's kind of intuitive because it's a more contaminated environment. But the thing is they can root, they can dig, you know, and they bring all those. It's that prediction error that the brain makes. It's like, whoops, let me pay attention. Yeah, exactly. And you activate the whole brain. It becomes a fully motor kind of behavior. And, you know, yeah, I mean, the idea is, again, you engage more parts of the brain and that facilitates everything. So we've noticed that a long time ago that performance goes up. But again, publishing those papers is really hard into the biomedical stuff because they will say, well, we only have contamination. Well, anyway, going back to the sign tracking. So some of the conversations I was having with some colleagues about, for instance, the efficacy of the clicker, for instance, on its own, or occasionally on its own, I should say. Not I'm saying that it should always keep its strength on its own after being paired. But we realized after a while that it was a huge literature in rats about sign trackers versus goal trackers. And this is the idea that every rat out there falls into one of three categories, actually. Sign trackers, goal trackers, and the ones somewhere in between. And the idea, actually, interestingly, is that sign trackers, they're sign tracking rats, are much more likely to pay attention to conditioned stimuli that are around them in the environment. They're also the ones that get like it's a genetic trait for addiction, correct? Exactly. And that's where that literature is going now with the theories by Barrage, for instance. I'm a big fan of Barrage, as some people know.

But there's others, too, Tomy, humans, etc. It's demonstrated in humans, actually, that it's a big predictor, actually, of addiction. People that are sign trackers, the theory says, are more likely to become addicted and for longer as well. So I think there are some breed differences there as well. I think, for instance, that Border Collies are extreme sign trackers. I would actually, I would guess that Malinois are as well. And I think this is one of the factors in all of this, is that I think a lot of the conversations I'm having with some people where there's a disagreement on, for instance, the use of the conditioned reinforcers actually come partially from their experience with their own breed. And I have no problem seeing that some breeds would be much more goal tracking than sign tracking, for instance. And so that's one of the things that we're doing in the next year, actually, is explore this at the individual and breed level as well. Wow, that would be so interesting. Yeah. And it would make sense. It would actually be a full convergence with both the human and the rat literature on this, actually. It seems to me that it just makes sense. It's, you know, and a lot of the old stuff of Ahrens and Shoemaker that show that some breeds like Border Collies, for instance, and I believe Malinois as well, are higher in dopamine levels in terms of basal activity of the dopaminergic system, would go actually towards this idea that these would have to be sign trackers based on the neuroscience that we know about dopaminergic activity and sign tracking. So it's an intriguing idea, but I think it's one that makes a lot of sense at the end of the day that we are thinking basically is where I'm going with this of assessing our dogs here now, not just for their nose, their trainability, their motivation, but also trying to determine if they're sign trackers or goal trackers. Oh, for sure. Yeah, 100%. I mean, if you actually understand that there are two different ways to look at things and find out which one, changing everything, right? And what about the duration of sessions? What is an average session for, like a seasoned dog, like a normal, just a session? How long would it normally go for? Yeah, so we don't really push it here. I mean, so we will have a dog come two or three times a week to the lab. No more than that. We've noticed that sometimes, even for some dogs, three times a week is pushing it a little bit. I think it goes towards the repetitiveness of the lab tasks, at least they have to do here.

But now, as I say that, some dogs could probably come every day and they would be fine. But anyway, so on average is two to three times. And the sessions during a day are typically three, about two to three, sometimes four sessions in a period of about two hours. And each session has 10 trials. So it's not a lot, but, you know, a dog will come here for typically two hours. And depending on how well they're doing that day, how motivated they are, they will do between, yeah, two and four sessions, usually on average, I think three. And like I said, it's only 10 trials. So they take about 10 minutes each. So they spend actually more time socializing with the people in the lab and having a nap under the table or having a little walk around campus than they actually do working. Which, again, you know, especially since these are not competition dogs, they're pets, right, is what has been optimal for us so far. Now, again, we've had some border collies here that the whole time that the students are actually, you know, having a tea or something between sessions, they're at the door of the experimental room just waiting. Can I come in? Yeah, I'm ready. Let's go. Let's go. Let's go. Let's go back to work. Obviously, we have those dogs. But, yeah. So we don't push it. We're not in a rush. And we certainly don't want to lose our dogs. I mean, we've seen this, unfortunately, more often, especially with the lab stuff, not the field stuff. The field stuff, typically, you can't stop them. But with the lab stuff, I would say about half of our dogs with time will eventually kind of let you know, you know what? I've been doing this for a year or two now. It's kind of boring. It's always the same thing. So then we will get them into a different experiment. So it's a different setup or it's a different way of doing things or it's a different order. Or sometimes they are excused. Right, right, right. You know, of course. I mean, boredom and habituation, this is, you know, the repetitiveness of doing things. It's for us and dogs and rats and everybody probably to some extent. I mean, yeah. Keeping that type of motivation that it's a... I think we don't have enough. At least I don't know of enough of research done on specifically maintaining, like just preventing boredom. It's with everything we do. Like this is probably the... Even when you have talent for something, you still have a... Depending how things are presented and how things move forward. You can hit that point of, okay, I'm just done with this. We've had dogs that we gave a break to for a year, for instance. And that sometimes work.

You know, it is funny because when they come back after a year of taking a break, we call it a sabbatical. They often show up at the lab at the door, like super excited. Oh, I haven't seen you for a year. This is amazing. And sometimes we can get them back for another year of, you know, being part of a few experiments. But yeah, that's the unpredictable part of all of this. And this is why, unfortunately, I'm moving away now from the biomedical stuff. Laura actually will be my last student on biomedical projects. Because what's happening now, unfortunately, for a lot of the diagnostic or alert work, the electronic nose companies are catching up on the dogs. Really? And their point is that, and they're right about this, is you don't have to worry about motivation with a machine. They'll do their work, push the button, and they'll do it. I mean, they can fail the same way that dogs will fail as well occasionally. But I don't give to the biomedical stuff a lot of life, maybe another decade. Ten years from now, I think that electronic noses will do the COVID detection in airports, for instance. We will detect blood glucose levels from breath easily. We'll detect anxiety also probably, or impending panic attacks from breath and stuff like that. I think we're probably just about max decade away from this. So that's why I'm going back to the wildlife conservation stuff. I'm an ethologist at heart anyway. I work with all those other species, birds, reptiles, et cetera. So for me, the dog is a research assistant in the field. And again, I've never seen a dog losing motivation in the field. It's a very lab-specific kind of thing that happens, you know, because the lab is so predictable, right? So I wouldn't say necessarily boring, but yeah, it's predictable. It's always the same thing. You make me sniff those vials that I've sniffed a hundred times. You know, it's always the same order for the last three months. Do you ever go, do you ever get the dogs in the lab, but actually don't do detection, but just have some other activities to where they're like, well, this is... Yeah, I know where you're going with this. Yeah, we, yes, we, it's not very practical to do, but we've done it with some dogs. Sometimes it takes time away from other dogs to come in to do work, but we try at least to make the experience of coming to the lab interesting. But you know what? This is dangerous because we decided with some of the dogs that were losing a little bit of motivation that maybe we would have playtime with them at some time, right?

But then what happens is the playtime that we had overshadowed the work, and it was even harder to get them into the experimental room. And even if we had associated the experimental room with play, in fact, that was even worse because they were like, oh, we're here to play. No, no, no, we're here to work. Now, again, with some dogs, work is play, so there's no problem. But it's that 15% or 20% that sometimes realize that, you know what? No, this is really work. And we never push the dogs. I mean, from an ethics point of view, you know, when we have a dog that loses motivation, we tell the owners and we say, you know what? Maybe it's time for a break or, you know, your dog is not enjoying this anymore, so there's no point. We stop. So we never force the dogs to it. But we'll always try a few things. But we've realized that often it backfires to try to make things, you know, to associate the lab with really more fun. Mm-hmm. Like just a little bit too higher of arousal and cannot really get into. Yeah. So we have to be very careful with that. It has worked with some dogs, but with others, it made it much worse where they would say, sure, let's play. That's a great play stuff. I see my friends once or twice a week and we play. But yeah, they would not even walk into the experimental room. And it probably also will depend on the level of just understanding, as we mentioned very briefly in the beginning of the conversation of it's not just play. It doesn't just mean let's just go crazy. There's still got to be some rules. There's still got to be some goals and objectives and boundaries. And I think that will play a role. And also if, like we have, it's a, I mean, it's a, you know, problem with competition dogs as well. Once you start from a little puppy and you just keep working on something that is still the same. And what, at least I do a lot of times is I always keep in mind that boredom will be an enemy. Eventually it will become a problem. And I would go sometimes to places and the dog is just like with your dogs coming to the lab, my dog going to like a stadium. And the moment they see the stadium, it's like, okay, this is showtime. And I would go some quite, actually more often, I would go places to a, you know, and other trainers will be like, oh, let's watch him how he trains. Pick up on some things, right? And, but I don't train. I just kind of ask maybe one or two things from my dog. And then we just go into just having fun for the next 10 minutes and walk away. And the trainers, strangely, they're like, oh, he just didn't want to show us how he trains.

Not understanding that all I'm doing is making sure that my dog goes every time with that excitement of, wait, this is going to be a cool thing. It's not a chore. It's not a, yes, there will be, whenever there is an element, whenever we say, okay, let's interrupt the play and let's do something. Yes, there is some compliance. There is some obligation, but I want you to want it. Otherwise there is no point. Who cares if you can do it, but you're pushed into it. That's not the point of dog training. Yeah, and, you know, with the dogs that we were talking about at the beginning that are not, you know, sensitive to food as a reinforcer and work better for play. When we do that between trials, that has to be very disciplined. So, for instance, we have a dog, you know, gets the click. That was a good response. Then we have the ball. We throw the ball once. Brings the ball. It goes back into the bag. And now it's time for the next trial. Yeah. The students have a really hard time keeping that discipline because you basically have to train the dog to understand that playabout is going to be 10 seconds at most. And, you know, it's one throw of the ball or the Kong and that's it. And then you go back to work if you want to continue. And that's easier said than done. I have to say play. And I think Ken Ramirez said that before. At least I was told. Play can be problematic that way. You know, to get a dog to get off the play mode, you really have to be clear that, you know, there's a temporal boundary to this. You know, that playabout is going to be 10 seconds. Then we're back to work. And but so far with the dogs that respond to play, most of them we've managed to train to understand that. You know, the playabout is short. If you want to play more, go to the next trial and, you know, give the right response and you'll be all good. So with play, like my I'm probably one of the biggest proponents on play as a motivation. Like I can never say never, but I really do not use food as a reinforcer. My all of my training is play based. And I divide it on like we have play as an activity. And these are just terms that I use for for my own people when I teach. When we have play as an activity, but we have play as a game, which means it's a it's a structure. There is an objective like like just like when you play basketball there. You have to have the hoops. You have to have the and you don't just grab the ball and run. There is rules. There is penalty. There's it's a very structured game. And if I teach the dog, assuming again how we were talking about genetic predispositions, what what do you like?

And I if I can inject some of those genetic instincts or whatever we want to call it into the game. But we still have an objective. We have rules. We have penalties for breaking rules. But it's a it's a very structured game. And I can show the dog that they're really good at that game. But not so good to where they feel that they are so superior that there is no challenge anymore. There is no competition anymore. There is nothing that, you know, but the game is always. This is the interesting part, right? Any time, any game that we want to go and watch, it's always we don't know exactly. And that's where I like in a very early age with really with every dog that I do. I am not preoccupied with teaching a sit and down and, you know, kind of all the classical looting things. And and. And I just go into play and teaching a game because I find that if I teach the dog a game. Before they even know sit. We. In that struggle for me to explain what the game is about and that they're good, but not so superior. When that happens, I actually have so much going on from that point as far as like I've accomplished a different level of cooperation. Like, I mean, you know, like you can we can look at just wolves or any dogs, you know, you can have the strong dog, the weak little dog. But if both want to play, the bigger dog will never just slam him to the ground and get the stick and go away because then that's the end of the game. He has to level up and make the game interesting. And. I spend that. That's kind of. It always been my go to. Way of maintaining motivation and. And then. Talking about behavior problems, talking about all sorts of things, you know, like if if that dog understands how to play a certain game and he knows the rules and he knows that he's good at it. I always give these analogies of, you know, like you can have a little kid from from Europe that plays soccer and we can throw him in Brazil. All by himself. In the soccer field. And within five minutes. He's accepted. And they are doing something together. And and you can totally manipulate control and self confidence through play, which you cannot really do. I mean, you could, but it's almost very. Artificial if you try to do it in a different way. I mean, that's kind of what games are about. It's all about if you're good, your self confidence is if if you can outsmart. Then you have strategies. Then you have to understand that you have to restrict yourself from breaking rules and so on. There's just so much more. I don't even know how I went there right now and why I went there. But I really don't know what I was talking about.

So the Panksep or actually Barrage theory about about using play also would suggest that play is indeed a very potentially very strong reinforcer. Because it produces gets you to produce endorphins the same way that food would, for instance. But again, the key is to get the animal back in anticipation. This is why when you say there are rules for us, it's the stopping rule basically is that, okay, that playabout is done. But I mean, you have to realize, too, that the stuff we do in the in the lab, when we do a session of, let's say, 10 trials for a scent detection, it's pretty quick, right? So, I mean, I don't have a lot of time to playing. It distracts the dog, too. And so in the field, I'm much more likely to use play. There's more degree of freedom. There's more space. There's et cetera, et cetera. But, you know, going back to the motivation part of things, play is very powerful for dogs to respond to it as a reinforcer. There's no doubt about it. And another thing that we've noticed over the years is based on the reaction of the students often when a dog suddenly has a decrease in performance. And this is something that happens in the lab. The first instinct of the students is to go back. So they go right away into the baby step kind of thinking. And if it's too hard, just make it easier. And then the dog will come back. It almost never works. What usually works is what I tell them to do, which is, no, make the task harder. And usually it's amazing. And the students never believe it until they see it. That's when the performance comes back. That goes towards that boredom comment that you made earlier. I think a lot of the time we underestimate our dogs and we get them to do things that, you know, they're ready to go further. Yeah. It works a lot with sun detection for us. I usually tell the student, yeah, this is too easy. And the game is over then, right? If it's, you're always right. And you know that you'll have this 90, 95, 100% of the time. Okay, you get the kibble. But there's that other dimension of the task that is completely gone. And so we, then we dilute the odor or we do something to make the task harder. And then suddenly the performance goes back up. And you also see that the dog is more engaged. The tail's going a little bit more. The sniffing becomes even more, you know, and there you go. Yes. And it's not so much, it's not so, it doesn't necessarily even need to be harder task. As long as we trick the brain to where it's like, oh, this is not what I know. You change, well, yeah, exactly. Changing the game. And again, this for a researcher is a nightmare.

Because if you try to write a paper where you say we have to do this with this dog and, you know, and then they will be all over you on this. We do it anyway. But, but yeah, no, absolutely. It's the understanding, I think, that, you know, if you go back to the Rescol and Wagner model, actually, it's, it's indeed this idea that, and as we know, actually, it goes beyond Rescol and Wagner. It's not just the, the, the US. It's also the CS that's important in all of this, like Pearson Hall and all of the others, Macintosh as well. A lot of these great models that were developed in the 70s and 80s basically showed that the worst thing you can do is make things way too predictable. Yes. Yes. So we, we go back to what I was saying about, do you need to always pair the CS with the US? A lot of people actually throw at me Rescol and Wagner to make their point. And I always find that funny because if you actually look carefully at what happened after Rescol and Wagner, first of all, Rescol himself admitted that his model was not very predictive of everything in classical conditioning. It had some flaws. It changed his theory, changed his theory, actually, or, you know, help to identify some of the issues with it. And then as Wagner of Rescol and Wagner that they've loved this whole model. And my favorite is probably Pierce and all that they've loved the model that actually showed that interestingly, that predictiveness is precisely the problem. And they have this great experiment, I think it's Kay and Pierce actually, showing that actually a 50% pairing of the CS with the US is actually working better than a 100% pairing. So going back to this idea that you always need to pair the CS with the US on a one-on-one ratio, actually the answer is according to that model, no. And it goes back to this idea that anything even in classical conditioning is going to eventually get predictable and predictable is boring. Right. It's basically what those models are eventually saying. And that's kind of where the, I don't know why, I mean, you probably have noticed this too, but the dopamine word is big in the last few years in dog training. And a lot of times it's, I don't think people even understand what it is about and how to do. And I'm almost afraid to even talk to you about it because like, you know, so you can put it so much better than me. But, I mean, it's always the search for something new. That's what really triggers dopamine. It's the moment you have it, it's done. Game is over. Like you're, just like if I win a world champion tomorrow, how long it's going to last? A day? A week? A week?

And then there's gotta be a new chase. And that's what dopamine does. It strives on chasing that new something that you cannot. Something that it's... Yep. And that's these incentive motivation models I was talking about. It's exactly where neuroscience is at right now. um i you know i'm a little bit responsible for putting that dopamine thing out there and i kind of regret the way that i initially talked about it i think that comes from sparks 2014 or one of those conferences and um um you know if if i had to go back i would introduce the concept differently the mistake i made at the time was to try to explain this from a neuroscience perspective and what i've been doing in the last year or two is to go back to the learning theory models that were making exactly the same point without the neuroscience um and uh yeah it's it's the last three or four presentations uh or webinars i've given actually and that it those models are either forgotten or just not known and i don't understand because there's even some young faculty that that engage in this debate that don't seem to know about these models um but anyway they they are they are quite powerful because again it's this idea that you can teach you can condition motivation which is very powerful if you think about it wow and it goes exactly towards what you're saying which is basically you're not learning if you're bored and you're not learning if you're not anticipating something that's potentially exciting uh doesn't matter if it's food play or something else or even praise right and even i mean to go even further the brain doesn't even ultimately it doesn't even make difference in in is a negative is a positive reinforcement it's it's it has to intrigue the brain that that's really the bottom line somehow yeah it's you know i you have to be a little bit careful about you know like how if if you start using aversives and and stuff like that but again i mean on that debate which is another one i i think we've lost a little bit our common sense here the way we discuss about these things and all black or white but um but in a sense if you do work with the anticipation of just of reward anticipation of a reinforcer you don't even have to think about using anything um aversive or punishers um but again if you establish just a very very boring and extremely predictable link between uh um you know the the the the the condition of the reinforcer and and uh the primary or or you know whatever how you reinforce your animal then then yeah then then you need to be

much more creative and it's this idea that how do you engage that brain again that's not understood and we go back to we go back to intermittent reinforcement the idea also that if you don't know that you will get reinforced it doesn't mean that it's punishment um right if you don't get reinforced it doesn't mean that it's punishment it means actually for a dog that's well engaged in the task what's next let's get to the next trial so i can get a chance to get reinforced you know people often tell me aren't the dogs that uh are not reinforced every time getting angry at you barking biting that kind of stuff no and the ones that do usually it's only because the student went a bit too fast again going back to this idea that you can't go super extreme on this uh and if they do it's because they haven't developed yet what i would call this you know incentive motivation uh they haven't developed yet this intrinsic motivation and that's because you didn't do it right or or you went too fast i know i know i know that you like like by like like who you are is uh from from everything that i've read and and by the way there is the way i i got introduced to you and then once i read this one article then i kind of went into search and and listened to a bunch of podcasts and some presentations and so on but there was um what was it there was that one 50 shades of gray misuse of misunderstanding and misinformation of the concept of dominance and punishment and i am like if there is three things that i am really about that's play um all all reinforcement and all varieties of punishment and genetics like i've been breeding malinos for since 89 and i've done so many and like pretty much any little research that came out i was able to just do it myself with my own dogs to see it's like you know like the military neuro like early like all the little tests even including um back to going back to was it faffenberg or who was in the saying how they did selection with the guide dogs and 50 percent increase the rate of success and then i ended up working there and i'm like wow it's still 50 percent chance of succession nothing has changed regardless of of all the tests that they the volunteers still are doing and these kind of things are always puzzling me and very very interesting about but what i like the i i i know you are big on force well force free stress free fear free all all that concept right and i i really want to get a little bit to hear what your opinion is on like what i'm thinking because we just just kind of like how you talk in your some of your articles like oh yeah we know punishment

works i mean but do we need to use it is it ethically ethically and morally and so on um and then we talk about negative reinforcement and i always think and i always when i use like i like for example if we play let's say we play that game of hot hands right that would be a game of negative reinforcement and there is a very cool incentive and when done properly there is not necessarily fear this actually i mean we play it all the time as kids uh at least back in europe that was a you know one game that kids play even even even even with a two-year-old you're playing such a game and there is we know and for some reason we avoid talking about it but we know that avoiding something unpleasant is more important than getting something pleasant we know that it's fundamentally at least in our planet if it's not the most fundamental law biological that you avoid something unpleasant and you approach something pleasant so this is this is here it's so deeply programmed billions of years to the single cell organism correct like it's i don't think we need to learn it i think this is something that we we know very well how to respond now can it be done wrong no question about that but it's here and it in my opinion it's here to stay and it can be it can be dangerous it can be dangerous or it can be highly beneficial and i always hope that one day i can make i can like i could send you videos for example of of dogs that i train that i have trained from puppies in a dog train through positive reinforcement and a dog train through negative reinforcement to a year old now knowing a bunch of commands and have super cool motivation and both dogs will be off leash off training colors of anything there will be no difference in the way they will perform or in the way they feel uh uh their emotional states they will be off training for them because it's they both have the willingness to play they both have the willingness and the confidence to to succeed i i guess the point i'm trying to make is that there there is people will do that just like dogs will do that i mean dogs play with each other right when when when we like at least my understanding when i look at two dogs kind of chasing and biting each other to me that is a negative reinforcement game to tour a very skillful and very it's actually a lot of fun game because it's kind of like playing some martial arts in a way do you do can you see it like this two dogs playing as a a negative reinforcement kind of game or or my thinking about this is that i think we in this conversation of the quadrant and uh

you know this over emphasis on the positive reinforcement uh i i think we we derail completely sometime in the 90s maybe early 2000 about this by by first of all misinterpreting a little bit what the quadrant is saying which is basically that in a way any life life is about the balance between all all four the contrast often between all four is often what's important but the ethics of modern dog training is focusing on only positive reinforcement which by the way is not actually uh the way that we conceptualized it as humans necessarily uh appetitive or positive i mean positive in this sense here that it's seen positively by the dogs we decided it is right we don't ask the dog because you know the quadrant what we forget is that it's all about the perception of the dog of what they think is uh punishment or uh or reinforcement so i've seen dog trainers um in french we say acharnement that uh that that can't give up with the clicker or whatever you know and they persist and they persist trying uh with positive reinforcement when actually just one one no right would have stopped the behavior but there's this allergy now of uh this aversion in dog trainers to ever say no to a dog no you don't say no to a dog it's punishment well wait a minute actually is it i mean first of all you could make the argument is condition inhibition you don't actually have to go this far uh it's just a way of saying you're not going to get reinforced right now so in a sense it can be a non-reward marker in some cases right that doesn't remove uh that doesn't remove the the whole controversy around it but you know one of the point i was making in that article that you mentioned 51 shades of gray was to say that well you know we don't have to see a discriminative stimulus or a non-reward marker as uh as something bad first of all it's not necessarily a condition punisher but the way that people define saying no to a dog they define it as a condition punisher that's not what i was talking about um okay so there's a lot lot around there here that you know again it's what is aversive to a dog that's the first question that we have to ask ourselves you know if we go with uh uh negative uh reinforcement it depends what is the initial state of the dog if if the the state that you you you get them out of is extremely aversive then i'm against correct uh negative uh reinforcement but uh if it's just a question that the dog likes to be off the leash when they get to a certain area where it's off leash and that's seen as a reinforcement or great that's that's no big deal i mean you still have to walk from

your house to that location anyway with the dog on leash um what i'm saying here is that actually i think we we apply negative reinforcement and punishment daily with our dogs a lot more than we think and we drive ourselves nuts in trying to avoid it at all costs in training when we still use it more than we think we're not even aware of it yes part and part of that problem i think also is another thing i talk about a lot in terms of the psychology and the neuroscience of this concept which is resilience it's this idea that we're in a culture right now where we've accepted this strange idea that with our kids and our pets we should protect them from any form of stress of life the problem with that is that anybody in psychology and neuroscience would tell you that that's not how you train resilience resilience is not something that you will develop if you're always kept away from stress or anxiety so that means that stress immunization is something that like the immune system is developed through being challenged so i'm not saying you should use punishment to train your dogs or any kind of reversives or anything stressful i'm saying this idea that we must at all costs only give them things that they they see as positive or appetitive is actually potentially misguided i think that a dog needs to experience the normal stress of life not extreme the normal stress of life um so that's you know that's a lot here conceptually to think about but i think we we are kind of in uh the middle of a pendulum movement from you know the 80s where punishment was used quite a bit uh shock collars and everything to this other extreme right now where no you must not at all uh ever say no to your dog or or indicate that they've made the wrong a wrong choice or decision or whatever and i think eventually we'll come back somewhere in in in in the middle um i have a tendency to think of drug training based on the kind of relation i want to establish with my dog and i parallel this with the kind of relationship i want to develop with my kids so i'd rather have a good positive relationship with my kids based yeah mostly on reinforcement but i would delude myself to think i never say no to my kids obviously i do um um and you know the danger here is is how are you translating or or moving this into dog training um so i'm not pro punishment it's not what i'm saying but at the same time i do not i'm not against informing a dog politely nicely of a wrong decision in other words you know in the discrimination tasks that

we do here with biomedical samples when the dog gives the right response they hear the click they get the kibble or a short play bout but if they get it wrong they get thank you right and that's simply a way of saying you failed this trial so walk out of the room and then let's go for the next one and it takes five seconds yeah which which is what we what i i talk about normally when i teach uh it's understanding that progression is reinforcing and knowing how to manipulate progression of the event and if you if if you can show that that progression can stop it's super powerful yeah and the thing we never see a dog i mean sometimes they'll go up oh you could see this oh i missed that one but what they do then is they run out of the room because they can't wait to get into the next trial right uh and again it takes about five seconds by the time the door is open again and they can go to the next trial that the the assistant just has the one of the two or three assistants just needs to have the time to to switch the the you know play with their random order of the uh the targets uh in the lineup or the two afc that we do whatever but that's the idea is that it's uh it doesn't have to be uh seen as a punishing event or an aversive one it can actually contribute more to okay okay you have another chance now let's you you didn't get this one let's let's go and get the next one right you keep it positive um some people say why don't you redirect the dog towards the right answer well we do this initially in the initial training you say oh no but look that's that's the one look look look at that sniff oh yeah there you go good dog and then we you know we do a what we call an error correction basically we say okay you got that one but after a while you have to stop doing this because then if you if you if you keep doing this they will know that any trial that they do they will eventually have the right answer and it stopped working and they don't make any effort anymore so they do have to experience that whoops that wasn't it no thank you let's try again but you would not believe the number of dark trainers that have told me no you can't do that that's not ethical that's that's just aversive training like well yes it becomes very very strange like it's a it's a difficult place where we are i i i always hope that somehow and i don't put myself because i'm i i i i don't even know where i belong i i just train dogs and whatever whomever is in front of me bottom line is that that dog is emotionally stable

and motivated to do things with me and when you said earlier by the way that i said punishment works let's be clear for people that did not read 51 shades of gray what i was saying is it works but potentially with consequences i mean it's right so when people say punishment doesn't work no no no it actually works but but there are side effects of it that you don't want if you want that stable uh positive relationship with your dog right or with your children as i use that analogy sometimes i i don't want to be the the naggy parents that always says no and you know no i mean there's ways to build a constructive relationship based on positive reinforcement mostly um that will actually work but you know it's it's a weird conversation because i i often talk to dog trainers where i realize that that even we have weird perceptions of what we think punishment is i'll give you an example for instance they'll say they'll say uh well when the dog gives the wrong answer instead of saying no or no thank you why don't you um why don't you just ignore this to me is fantastic that they don't even see that actually ignoring is often worse the only time that i got a dog barking at me or even sometimes trying to to snap at me was when i would not give it the information so they seek information they'd rather know that no that was wrong than me just doing as if nothing happened and get them out of the room they go like what the heck what's going on here so i don't see it as punishment when you say or a non-reward marker even although i i guess that word could mean that in a sense i see it as giving feedback you say yes good dog yay fantastic it was great or no thank you thank you no that was not it let's try again right and you can keep it positive it's just a way of saying all right you're not getting reinforced on that trial no big deal let's go to the next one right away so but if i play devil's advocate here emotionally speaking the dog and that kind of goes back to what you just said very just just a while back a lot of trainers that are are considering themselves for free but they fail to teach something correctly and there is a struggle and there is no physical uh punishment used that's not a problem but mentally and and very visible visible you see that the dog is in a bad emotional state because he he or she understands that they need to do something they are trying it's not leading anywhere and now they're becoming kind of suppressed and experiencing very similar event on on a uh as far as the brain is

concerned like if i have uh uh one of uh kind of simple example that i talk about um i'm obviously with my my accent you know i'm from eastern europe so let's say we we go we go back to bulgaria and we go to a gas station and i have this really funky car and i say simon i'm gonna give you 5000 euros if you can fill it up in two minutes and now you have to find where the where the tank is right and you have completely you're super motivated to find it now i can let you struggle or i can say no that's a wrong idea you're going to the spare tire there will be element of aversion that will actually at least if it happens to me i will be grateful because you are helping me to get unstuck from somewhere instead of leaving me alone to figure it out does that make sense i i used a very similar analogy uh uh when i was at a conference in erice a toadog i think it was called where i said look uh the the way i see it is this um you you get in a cab and uh you're telling the taxi driver that actually doesn't know the city at all to go to the airport so tell me which one will get faster to the airport the situation where you say only yes when the taxi driver makes the right turn or the one where you say yes or no before they engage in a wrong turn uh or after the wrong turn uh i think it's clear that the one where you indicated you gave the feedback positive or negative will get to the airport faster than the one where you just give one type of feedback um and in other words yes it's not getting information that's a cognitive theory obviously is is potentially extremely extremely frustrating there's another dimension to this that a lot of dog trainers don't realize is if you look at the work of psychologists that work with kids so my wife works with uh with children in hospital setting also in kind of uh procedural issues where they have to go through some very difficult procedures sometimes that come with pain and those kids will uh you know kick the nurses and spit at the doctors and that kind of stuff and and uh her job as a psychologist is to walk in and try to change that situation she points out to me that people that work in um pediatric health psychology actually know that sometimes uh persisting with positive reinforcement in those situations will take extremely long will be extremely frustrating and may not work at all and then the ethics of this is after an hour or two of trying to get the kid to accept the procedure they're still super stressed and they end up having to strap them down when actually if you had just walked into the room let them play with an ipad for five minutes and then

say okay if you want to play more with this you will have to do the procedure and then if they don't comply you walk out with the ipad which is basically negative punishment correct but then the procedure happens ethically there's no doubt that the second scenario is better than the first you solve the problem in maybe 10 minutes instead of spending two hours persisting with positive reinforcement to no avail including the stress and the resistance and the back and forth so this idea that positive reinforcement reinforcement reinforcement is always non-aversive is wrong correct it it's not necessarily true especially not in the way that it may prolong the period of stress and sometimes a uh yeah a an aversive relatively aversive intervention is ethically desirable because you reduce that stress to again five or ten minutes as opposed to an hour or more now again when when i say this i sound like i'm pro punishment and i'm not i'm not but we have to realize that there are some situations where uh we have to think about the ethics of it i mean i think we always need to think of the ethics and you know what is uh kind of in a lot of ways it's unfortunate and sad that that word is packed with so much confusion and and immediately brings uh uh just it's it's it's not a like it's very difficult to talk about it um when when we try to talk about punishment especially in dog training and and with kids just as you said this is a very very big problem um and somewhere it went wrong and i don't think anybody knows exactly where it went wrong well you know part of the being wrong look if you if you really believe in this concept that you never say no to a dog that means you probably believe to the idea or often anyway you never say no to a kid well look at what kind of generation we've created we are dealing with that generation now that was never told no for anything and then we complain that they're entitled well where do you think that's coming from obviously they are entitled and they feel entitled because we never told them no uh so i mean it's again life is not that simple life is not all rosy it's not just white or black uh it's not even black and white it's shades of gray and that's the whole point uh it's complicated you may want to avoid the extremes but the idea is that yeah it's not just one dimensional there's many different factors in all of this it's not as simple as some people want it to be unfortunately yes and there is i mean really ultimately there is cost and benefit at all the time and like like with

the example you gave me with your wife and the kids um i think one of the big difference that we don't like to talk about between dogs and kids it's super obvious and like we we can really say hey you're not gonna get the ipad unless like we can instruct it becomes a little bit more interesting the approach when somebody has a little well not a little but some serious uh uh mental uh disabilities then then the approach is not so straightforward because at that point negative punishment not necessarily makes sense sense the way uh um you and i can make sense of and when we bring it back to dogs for dogs i personally believe that they are way more easy to understand and respond to that simple biological fundamental fundamental thing of i will approach something that i like and i will avoid something that i don't and and somehow that went wrong in terms of dog trainers start to do stupid mean things to dogs to where it just you you either have to call somebody and take the dogs away or you have to walk away or or something because it's just it's heartbreaking to to even listen to hear it what's going on but that's not like like we're just bouncing from one extreme to another hoping that there can be a solution and if we talk i mean if we talk about differential reinforcement ultimately differential reinforcement like just like if it's impossible which actually i want to talk to you about this i want to ask you uh um what punishment based training means means because i i never get to understand when somebody says that but i never get the chance to ask what that means but but let's say just because i think it's very difficult to teach new behaviors and try entirely through punishment it's just i don't see how but it can be also very difficult to suppress uh undesired old behavior in entirely through uh reinforcement of incompatible behaviors and if if reinforcement of the undesired behavior cannot be completely uh um eliminated then what do we do then then what's going to happen is likely it will continue to to be there and now we will have the the new behaviors the differential reinforcement behaviors and ultimately we will have a uh concurrent schedule of reinforcement with depending uh which behavior is available you know what i'm saying and so in in this situation some form of punishment see i i even hesitate to say it because it's just it's such a word that it's packed negatively but to me the best formula would be to to be able to suppress

the unwanted behavior and use all the cool differential reinforcement to to to redirect well you know and so i i don't have situations where i would go there in what i do but i here's the thing the the like i said earlier i think you know a huge part of the the problem is we we've uh in french we we talk about that but we we've gone crazy about the the concept punishment as well uh which is the idea that there is mild punishments and there's not so mild punishments obviously uh yeah i i get what you're saying i guess i'm not sure i i mean i don't i guess like my my thing is why do we not talk about like trainers scientists why do we not why can we not have the conversation yeah well i think that it's it's it's it's it's in part because we're dealing also here's the thing i think we're dealing with for instance people that use shock collars okay um this is not a method i like personally um in the 80s when i first started training uh i used it once and uh i i i realized it was not at all working and and back then by the way the the positive um training uh you know ideology didn't didn't really what was not around uh at all in fact i learned this from military people actually so anyway um but you know again i realized this is not the kind of relationship i want with my dog so i walked away from that completely um so i don't know i i i was going to go with something here i'm not sure i want to go there which is the idea that again a lot of people think they're not using ever punishment but they do actually in some ways uh and i was also going to make comments about the the you know what what again what kind of quote-unquote punishment you use and is it it actually punishment or is it extinction for instance right i mean there's different ways of stopping our behavior um it doesn't always require uh uh punishment per se there's conditioned inhibition that's you know purely classical conditioning there's anyway there's all kinds of different ways correct um but again i mean i think we just have to bring it back to the quality of the relationship that that that we have with our dogs and i think that's you know i i think one that is positive and by positive i mean uh you know happy uh there's no resentment there's no uh you know there's desire to play there's there's affection there is uh there's good communication this is what we should strive for basically um and anything that's aversive to a dog well it will happen anyway even with the you know they're all positive people uh but question is is to try to minimize that as much as we can um and

yeah yeah hey listen i kind of have to go that's fine we've been like i mean mine we did like a a marathon like i i truly appreciate like this was a very very cool conversation simon thank you like that was very very cool conversation i know like you just said it and now i realize man we've been talking for a long time yeah a little bit over two hours i think thank you so much it's fun i would not do it if it wasn't fun so i do it because it's fun it's played for me very nice thank you and have fun training tonight Thank you.