In this episode of K9 Conservationists, Kayla speaks with Tony Harvey about intelligent disobedience.
Review Highlight: JPinCB says “Kayla’s authentic, enthusiastic interest in all of her guests and topics, paired with her knowledge in a variety of scientific fields makes this podcast fun, interesting, and informative. If you have any interest in dogs, nature, travel, science, even pursuing your dreams, you will love this podcast.”
What is intelligent disobedience?
- When the dog goes directly against the handler’s cue in an effort to make the better or safer decision (ie. A guide dog not crossing the road when asked when there is a vehicle coming).
How do you teach this?
- This is taught by training dogs different behaviors for different contingencies
- A high level of reinforcement is given for these behaviors
How do you teach handlers to respond to this?
- Many sessions with and without the dog
- The handlers need to learn what each behavior feels like
- Are there specific dog traits that make this training easier/more successful? – From Taylor
- Do you factor in the need for intelligent disobedience when choosing a prospect? – From Megan
- For different types of dog personalities, how does he approach this kind of training, and some problem-solving examples? – From Janna
Links Mentioned in the Episode: Red/Green Light Paper
Kayla Fratt 0:09
Hello and welcome to the canine conservationists podcast where we’re positively obsessed with conservation detection dogs. Join us every week to discuss ecology, odor dynamics, dog behavior and everything in between. I’m your host, Kayla Fratt. And I run canine conservationists where I trained dogs to detect data for researchers, agencies and NGOs. Today, instead of a silence highlight, we had a review highlight, which is from JP and CB that says Kayla’s, authentic, enthusiastic interest of all her guests and topics paired with her knowledge and a variety of scientific fields, makes this podcast fun, interesting and informative. If you have any interest in dogs, nature, travel, science, and even pursuing your dreams, you will love this podcast. Thank you so much for the review. And if you haven’t reviewed the podcast yet, please hop on over to your favorite podcasting app and do so. So today I’m talking to Tony Harvey about Intelligent Disobedience. Welcome to the podcast. Tony, do you want to start out with telling us a little bit about your work and the dogs you share your life with?
Tony Harvey 1:06
Yeah, brilliant. Thanks for inviting me to this. Kayla, been really looking forward to it. So yeah, I’m Tony Harvey. My job role currently is National Dog Training lead as part of Guide Dogs in the UK. And I’ve been working in guide dogs now since about 2008. I think during that time, also worked at the New Zealand guide dogs as well. And epilepsy dogs in New Zealand. I’ve got my own dog. As you can see behind me, I know that’s no use for the podcast. But I’ve got my own little street dog from Thailand is behind me who’s busy on a Kong there.
Tony Harvey 1:38
And in my personal life as well, I live in this house, it’s in the country. So somehow I ended up with four goats and six chickens, couple of geese. So I do a lot of training with those as well purchase the clicker work with those doing sometimes similar stuff to what we do with the dogs, not getting them to guide people around the streets and things at the moment. But yeah, we trained them platforms in bits and pieces like that. And I guess my role in guide dogs. Previous to this has been very much hands on in developing how we train the dogs along with lots of other people. And we’ve got a training program within guide dogs, which has been devised by other people. And I’m one of those people just, I guess like to get hands on and try it out and see how it works in the real world as well. Really?
Kayla Fratt 2:20
Yeah, yeah, that’s great. So, as I said, we’re talking today about the concept of Intelligent Disobedience, which, you know, I’ve understood in the guide dog world as the example and you can correct me or expand on this in a moment of, you know, you’re asking your guide dog to move forward, and there’s a ladder overhanging, or there’s a truck or something and the dog doesn’t go forward. And I wanted to talk to you about this in the Search Dog context as well, because there are times where my dog may have caught an odor. And I want to call them back to me to give them a directional and perhaps in that situation, if I’m just giving them a directional cue, I’m actually preferred that the dog listen to the odor or, you know, obey the odor instead of me. But then there are other cases where if I see a rattlesnake, and I have an emergency down or emergency recall, I would like that cue to take precedence. And I think
Kayla Fratt 3:19
that’s one of the it’s one of those things that, you know, once you’re in the field, you may realize that you’ve got these different hierarchies of what you want your dog to listen to. And I would love to talk to you in the guide dog world about what that looks like as you train that concretely and intelligently for your dogs. And then potentially how we can apply that going forward in the field.
Tony Harvey 3:41
Yeah, it’s really interesting, because you mentioned Intelligent Disobedience to me a while ago now. And we started to talk about it and what it is and what it looks like. And the term as well, the Intelligent Disobedience, probably particularly the disobedience part of it, and what actually it means and what we’re talking about, and what the dogs are actually doing and what they’re thinking about and how they’re interpreting all of the environmental cues that they’re getting. And, yeah, as I thought about it, I kind of thought, what is it? And why is it called that, and it’s a term that’s floated around Guide Dogs for a long time. And I think in other fields as well, like you just mentioned the sent work and search and rescue. And I think almost if if we use it in its raw term as disobedience, we’re probably almost doing a disservice to our trainers and our dogs. Because if we think about all the work that people put into how they train the dogs and how they train the queues and how they maybe build up the queues as well, when they want them to ignore maybe another cue or another reinforce that is in the environment. But also for the dogs. If we think that they’re doing this almost higher cognitive work, and we we think that they’re doing by the training, then it might mean the next doc that we come across, we expect a really high level of them. So maybe we don’t put in the same antecedent arrangement to set them up for success like we might do. If we have a better understanding of exec What we’re doing when we’re training, what we refer to as Intelligent Disobedience. So one of the you gave us a small example that didn’t deal with the the ladder being on the street and what a dog does in that situation. So let’s think what is Intelligent Disobedience. So generally, if we give a forward cue to our dogs, it generally means just walk forward. And there’s going to be situations, when they can’t do that when there’s a ladder on the pavement, or there’s a car on the pavement, or there’s a hole in the floor, you know, all of those things. So if we gave them a forward and they did something that wasn’t just going forward, that’s what we turn that Intelligent Disobedience. I think in reality, probably what we’re talking about, is how we’ve trained a different set of contingencies for our dogs, so that they’re able to generalize to a vast range of different environments and situations that they come across. So for example, if the contingency is street heads completely clear, I give a forward cue, the dog just goes right forward cue nothing ahead, I just keep going. However, if we teach them another contingency, as well, when I give you a forward cue with the way a heads blocked, then I want you to do something different. And we do actually teach that we don’t just expect the dogs just to be able to work it out, we would teach them a whole range of different behaviors they could do in response to the contingency that they’re then faced with. So we’re given a forward, they can’t complete that behavior. So if let’s say it was the lad on the pavement, and the example you started with, what we would want then is for them to go to the side curb, so we want them to go to this side, so that they’re pointing out into the road. And from that point on the guider gonna could then go into the road for short time and then make their way back to the pavement. So that’s one of the contingencies that we might teach them as well. And there’s lots of other situations, whereby I don’t know if you’re at the train platform as well. So if there’s a big gap in front, and you don’t realize you’re on the train platform, you give a forward, you obviously don’t want your dog to go forward. But we will teach all these and we will build them in. And we always do it, everything we do, we build up gradually, as well. So when we first start teaching afford Q as well, we use platform. So it’s simply you walk a couple of meters to the platform in front of you. And then we would start to build up to bigger and bigger things. So when we start to introduce that concept of we want you to do something other than just go forward, we would do that in a very controlled environment. First, we might block the way to the platform first so that when the the handler gives the forward cue, the dog has to go to the left or have to go to the right first before it can pick up the straight line again. Does that kind of tie in with what you were thinking as well?
Kayla Fratt 7:30
Yeah, yeah, that does. And I think, you know, that example of like the forward cue and trying to figure out how to, as the trainer intelligently layer in and teach the dog what we want them rather than just kind of hoping that when you’re on the train platform, and you say forward, the dog isn’t going to be so quote unquote, obedient that it steps off the train track onto the, you know, electrified subway tracks or whatever.
Tony Harvey 7:56
Yeah, exactly. So the first thing is, you know, the first time we come across that situation in the environment, or when we’re doing it, we wouldn’t even give the Ford cue, we would teach them the alternative behavior that we want them to do first, and then we might start to build back in that Ford cue, but, you know, at a low level first, so we make sure we’re always setting them up for success, so they know exactly what to do. And we’re building up that history of reinforcement with those behaviors as well.
Kayla Fratt 8:19
Yeah, and in the guide dog world, this is important because you’re, you know, ultimately, the dog is being handled by someone who cannot see or cannot see well enough. And that’s the whole purpose of why you’ve got the dog and in the scent dog world, in a way, we’re doing this, we may want this because we as humans are nose blind. You know, it’s not that I can’t see the risk ahead of me, it’s that I don’t necessarily want to pull my dog off of odor. If I’m calling my dog to me, just to give him a drink of water. I don’t necessarily want to pull him out of a scent cone and off of odor for that. But if I’m calling him to me, because I don’t know, hear a moose or something. I may want him to actually listen. And that’s where I think so far in my career, I have kind of relied on the urgency and the tone of my voice to get the get the mission across or not. Yeah, get the mission across to the dog get the message across. Well, you know, I have a couple different versions of my recall queue. But I haven’t taught it in as clear of a way as I would really like yet. And I think that’s probably pretty typical in this field for us, you know, we like for my dogs, I’ll have like a bar Lee come. That’s kind of our casual recall. It works very well. They come very nicely to it. And then if they’re if they’re taking off after a deer or or there’s a car or something, obviously, like, just because of adrenaline, my voice changes.
Tony Harvey 9:53
So it’s really interesting. Yeah, go ahead. I think our dogs bail us out of a lot sometimes. because they work through all of sometimes inadequacies with our training as well, and they managed to work through it. I know what you really mean.
Kayla Fratt 10:08
Exactly, yeah. And I don’t want, you know, going forward, I think in an ideal world, I’m not relying on them being able to decipher different tones in my voice, I would like to actually teach them these contingencies intelligently. So and I know one of the things we talked about in kind of our pre interview is that, you know, maybe we are not sure about the term Intelligent Disobedience and we kind of threw around a couple others. Is there anything that you you’d like better as a phrase here? Yeah,
Tony Harvey 10:41
I’ve been thinking about what we might call it and things and as a phrase I hear, I can think of what we might label it really. Because what I was thinking as well, because it’s quite it’s a label, isn’t it? Intelligent Disobedience, we know that we’ve tried to break it down to what it is. And I was thinking, you know, sometimes the magic of science, you know, it’s actually in the science, and it’s in the knowledge, and it’s in the detail, sometimes. It’s not magic, we know that but it’s still pretty, like you’re saying, when we just try and give different tones of our voice. And that’s how we’re trying to train the dogs. But if we actually try and really train it perfectly, so we’re training that intelligence. That’s where the kind of the, you know, the real magic is the real sciences, isn’t it? So whether it’s separate to our training and needs a different term, I’m not really sure. Because I think it’s, you know, it’s something that we train all the time, isn’t it? So when we’re walking down the street as well, we have, you know, we refer to them as distractions, don’t we, but if we really look at what they are, they’re competing reinforcers. So we might have another dog that we’d rather our guide to walk straight past and straight to the curb for. We call it a distraction, because there’s another dog there, we don’t want to be distracted. But as far as the dogs are concerned, it might be really reinforcing to go and say hello to that dog and have that interaction there. And it’s all kind of the same thing. In that situation, we just want them to keep walking straight until they get to the next object, or next curb, when they find themselves in a situation or they can’t follow the exact cue they’ve been given. It’s exactly the same as that we just want them to follow the training that we’ve put in and we generalized, and we proved, and we’ve done all of that work to get the dog set up the dog for success as much as we can as well. And I was listening, I think you were talking to Dr. Susan Freeman, weren’t you a while ago, and she was talking about service dog industry as well. And it was very much sometimes you hear this term, don’t you? Well, they have to work in those conditions at some point. But the start, you know, our start position very much is not in those conditions, we don’t have to start in those conditions, we can start in a very different level from what we need to eventually get to, just so that we can really make sure we can get in that high level of reinforcement where we need it, we can really teach the dogs what we want them to do and really set them up for success. And another term that comes into the service dog industry quite often I’m sure it does in your world as well, is needing resilience. And again, it’s it’s another label, isn’t it? It’s like, what exactly is resilience? And think about what resilience is, I think it’s their ability to bounce back and have another go. And lots of the time, people would say, you know, in the past, they might say you need to build up resilience by making the mistake over and over again. So they build resilience to making that mistake. The approach that we take is we build up how to get it right all the time. So we build up that history of reinforcement. And that resilience comes from that ability to say, Well, I did it wrong that time. But I’ve got this huge history of reinforcement. Now I know exactly what to do to get it right. So I go and do that next time. And I think, you know, if we operationalize resilience, that’s kind of what it means to me anyway.
Kayla Fratt 13:32
Yeah, yeah, that absolutely makes sense. Yeah, I recently did an episode on kind of errors and mistakes and failure, because I got a question on Patreon from someone who asked about how do you teach your dogs to handle errors or mistakes? And I was kind of like, oh, I don’t I, you know, I really consciously teach them how to problem solve, and how to, you know, enjoy the game and, and work through things. And we build up their endurance and their stamina, and again, their creativity and their what’s the word kind of their ability to like, dig in and keep trying. But it’s, it’s not that I’m teaching them how to recover by introducing them to mistakes.
Tony Harvey 14:19
Yeah, you just bought errors in there as well, which is another really interesting point, because we talk about errorless learning a lot now. And I guess I’m guilty in the past of what we teach our dogs to stop at curbs, obviously. And I would have in the past said, you know, I think they only really understand it once they make the mistake a couple of times, and they realized the wrong as well. And I can’t remember the name of that pigeon study. But I think we mentioned it to each other where they have to pick a red light and how they introduce it by reducing the errors by introducing the gradient of light gradually. So that’s kind of my approach a lot more now. And that’s what we’re trying to guide dogs. It’s not think, Oh, well, you know, the dog needs to get it wrong. It’s like no, we need to teach it better. We need to build up that history of reinforcement more or We need to lower the criterion and you know, build up that criteria and build the difficulty into the work that they’re doing as well.
Kayla Fratt 15:07
Yeah, absolutely. So, you know, I think we’ve given a couple examples of how you go about teaching the dog this, you know, it’s almost it’s not so much that. Yeah, I don’t know if I like this Intelligent Disobedience term, you know, it’s kind of a cue hierarchy in some ways. But it’s also teaching them like, well, forward means forward in these contingencies. But in these contingencies forward actually means go around the obstacle, and then continue forward. Do you have any other examples of maybe maybe a third type of behavior that could fall into this category, just to give people another example of how, how you may problem solve this with a dog?
Tony Harvey 15:46
I guess I can, well, you mentioned search and rescue earlier, and I was talking to my search and rescue friend, just the other day to get an idea of how they might get that Intelligent Disobedience in. So similar to your sound work, but for them, it might be if they’re recording their dog. So there might be a distance as well. And if their dogs on, so they want to get their dog to search a new area, let’s say so they’re in the distance, and they will give the cue for new area. But if they’re on a scent, and they want their dog still to ignore that cue, and then come back to them, and indicate that they’re on the center of, you know, a person or something like that, and how they were building that Intelligent Disobedience to avoid to ignore that cue for search in a different area, and actually come back to the handle and indicate that they found somebody. So what they would do in that situation as they would have the, because when they train their dogs, they have somebody hiding with the dogs, and they would have not hide them with the dogs hiding for the dogs to find. And they would have a walkie talkie with them. So the walk, the person that’s being found, could indicate to the handler whether the dogs onto them or not. So then they would know how they might want to start mixing up the cues a bit. And the first start point for them would be really low criteria. So rather than just give a really clear cue for coming back, they might just be playing with their bag, or you know, looking at their phone or just building in, you know, slightly different body positions that they would do before, they would then start to show that queue for the right, so they do the recall, they wouldn’t show the queue for the right, you know, I want you to search over the right to start with, they might play with their phone first. But then they were gradually introduced that queue for the right. And then once they got it a lot stronger, and the dog really understood what’s been asked for it, they can build up on that obvious cue, which I guess is what you were doing with your voice really isn’t you’re kind of building up your voice a lot to get. So that sounds like they might run into that.
Kayla Fratt 17:32
This reminds me of a game that I did in back in a cent work class, I think I took online through fenzi Dog sports. And I think they called it an extreme proofing game. And what you were practicing is you know, as the dog is searching an area, the handler is like doing burpees or spinning in circles or, you know, doing all sorts of weird things. I mean, you build up to that, because a lot of dogs especially at first, if you so much as Yeah, pull your phone out of your pocket, scratch your head, rotate your shoulders, they’re whipping around to look at you and see what’s going on. And you get to the point where what you’re actually doing is you’re putting leash pressure and physical pressure on the dog to try to pull them off of odor, and then rewarding them for not, not listening to you. And I think where this starts getting really tricky for me and starts breaking my brain is it’s like, because I don’t have two different recall cues as far as like, you know, a whistle versus a voice or something where I would like to have. Because I think in an ideal world, I would have like an emergency stop and an emergency recall that Trump odor, and then everything else, odor should trump that. And I think you know, what I’m guessing I need to do is come up with a separate emergency stop and emergency recall cue so that I can really clearly teach this hierarchy to the dogs. Because we have done a lot of those proofing games, both of my dogs are really quite good at that. But that doesn’t then help again, in that situation where there’s a rattlesnake or something that I really need. You know, keeping my dog safe from wildlife is more important than finding a target.
Tony Harvey 19:10
I think another example of guide dog work as well as which you might ask the dogs to ignore cues. If you’re walking on the street and you want to go right you might start asking right, you know, before you get to that right turn as well. So when you first start introducing that, you’d obviously only do it when you’re right at the turn. But then as your dog got better at it, you might start to introduce it early so that if your dog turned early, it would walk you into a wall basically. So it’d be ignoring that cue until I guess the contingencies in place again, that you could turn right and you know, the space to go right. So yeah, I guess, in everything we do, because ultimately we’re handing over to a visually impaired handler. We’re trying to prove everything so that there could be an element of Intelligent Disobedience, if we’re gonna call it that for the sake of what we’re talking about here. We probably build it into a lot of the things that we do.
Kayla Fratt 19:55
Yeah, that’s, I mean, that’s exactly how I taught both of my dogs. We do some competitive skiing. shoring. So they have their directionals for that, and both of them, I taught them at a walk on sidewalks where, you know, I just said right as we’re making a 90 degree right turn, and then we build up to being able to do it at speed and where I yell it, you know, maybe 50 meters before we hit the branch in the, in the trail so that we can actually start taking those turns at, you know, at 20 miles an hour. Together, without them just, you know, the first time I the first time I tried that, before I kind of layered it in a little bit more, I keep using intelligently on the handler’s end as well as for the Intelligent Disobedience. But the first time I did it with my dog barley, he did take a 90 degree right into like four feet of powder in Colorado snow, you know, he took me literally. And you know, it wasn’t a big deal, we just had a tumble. And luckily, snow is soft. But that’s not the sort of thing that you’d want to experiment with, with traffic or a visually impaired person or anything like that.
Tony Harvey 21:01
Yeah, there’ll be other times as well, that we prepare our dogs for as well with our guide dog owners is that sometimes they’re given completely the wrong cues, or they’re told to go completely the wrong way. And it could be up a driveway or something like that, when they’re told to go left? And then we would really train the dogs, you know, no, your objective is to get to this curveball, ever. So we built in a lot of that as well for handler error, I guess at the end of day, yeah.
Kayla Fratt 21:23
Yeah. Wow. Yeah. So So is that you’re kind of teaching the dogs specifically to target towards like sidewalks and curbs so that they’re able to overtime learn to ignore things like driveways?
Tony Harvey 21:38
Yeah. So that’s primarily the main thing that we kind of teach the dogs who just, you know, keep walking straight, until you’re given a cue to do otherwise, or you eat something that you can’t get past or you reach, you know, an objective or curb or something like that, where you’re going to get your reinforcement. So that’s the the basis of what we’re actually teaching the guide dogs, and just to avoid all the obstacles on the way to get there as well, really. So in its simplest form, that’s what we’re teaching the guide dogs, avoid everything, stop when you get to an obstruction, or you get to a decision point. And I’ll just think you mentioned earlier about what we could call that Intelligent Disobedience. And when you said something, which made me think of a decision tree? I don’t know whether you’ve heard that term before. Yeah. Heard in tag teacher, I used tag teachers, well, quite often. And there was listening to them. I think it’s BJ Mumford, he teaches basketball and he does very similar to it, and he calls it a decision tree. So when one of his players would have the basketball, you know, the if there’s nothing in front, it might be shoot or dribble. But if there’s a player in front, it might be passed or something else. So depending on what’s in front of him at the time, you make this different decision. And that’s what he referred to as a decision tree as well, which I think works really well. I think there’s more, more what we’re doing, then we’re not asking the dogs to disobey us when you know, they’re doing some of the training that we’ve trained them to do. They’re making decisions based on what’s in front of them. So if we’re going to call it anything, I think the decision tree is quite nice.
Kayla Fratt 23:00
Yeah, I like that. And I think that gives the dogs the credit that they’re due for what we’re what we’re asking them to do.
Unknown Speaker 23:07
Hey, I’m Taylor, and I’m the handler for Kepler, a mini Aussie and training for muscle detection work before canine conservationist, I didn’t even know about all the possibilities with dogs and conservation. Now I’ve jumped feet first into the training, I wouldn’t have been able to without the support I gained from being a part of the podcast, Patreon. My favorite support comes from the group called I’ve been able to get alert training help and felt completely welcome. Even though I’m a complete novice to this kind of training. The group calls also helped guide my questions for my one on ones with Kayla, the information is invaluable and the community is kind. I hope to see you there.
Kayla Fratt 23:36
Yeah. So how do you how do you explain to you know, if someone was coming to you for their first guide dog? How do you help them understand? Or you know, and maybe it maybe it’s relatively simple, but just talking to them about, you know, if you give your dog a forward cue, and they don’t go forward, here’s what may be happening. And here’s how you can respond to support the dog? Or, you know, I would assume you don’t want anyone giving your dog a correction. And that, and that point.
Tony Harvey 24:04
Yeah. So this is the biggest part of what we do, I guess, because, you know, we have to work with dogs and people. And people have different skill sets, obviously, when it comes to training. But we now would spend a lot of time and I say now, because it probably in the past, you know, we might have just got the dog in the person together. And we try and coach through everything in one go while they’re in the environment. But now, you know, animal trainer would have a lot to do with his and how we work and we’ve got better at it. So now we would think about all the training that we do in the same sort of framework that we do do any training with any animal really. So it’s always about breaking it down setting up for success. And I always refer to refer to it as success orientated training rather than Error List that we can get as much reinforcement as possible, because it’s virtually impossible to get a list when you’re trying to set it up. So we would do loads of sessions with and without the dog. So for example, if we want dogs to ignore food on the pavement, don’t we but we also want the guide dog I need to know what that food on the pavement might feel like for their individual dogs that they’re working with. And, as I mentioned, we teach the dogs platforms as well. So they’ve got a very strong reinforcement of just walking up to the platform, and then getting reinforcement on the platform. So when we’re on a class situation, and bear in mind, we’ve already taught this to the dog in the incremental stages anyway, what we would do, we would have two platforms set out one at each end of a pavement, maybe 30 meters between them, maybe less, so that we could get the dog walking from one to the other. Before we use platforms, we couldn’t do that, because the dogs would get bored and switched off. But now we’ve got platforms, it’s very exciting, so they love doing it. And then what we would start to do is we would start to have food on the road or on the pavement at a certain distance. And we would talk through exactly what was happening for that particular guide, we’re gonna, and we’re teaching them what to look out for what they need to fill through the handle, whether they just need to take the lead and hold the lead in certain parts. And then we’ll get them to click and reinforce the platform. And then we will gradually move the food closer and closer and closer until they were walking over it. On top of that, we would teach them how that applies to the real world as well. So we would say, you know, if you’re going past a bakery or something, you’re more likely to have food on the pavement there. So you need to be more aware when you’re getting to these points. But also, you can be quite proactive, and you can already have the lead in your hand. So to support the dog to get past anything that you’re trying to get past. And those sorts of sessions that we do, we could do without the dog first. So we just have them with the handle. And then we we could walk them up and down first, before we even introduce the dog as well. And that’s kind of twofold, really, because what it does is it obviously builds up the client’s confidence, but it builds up their motor skills, so that when they actually do handle the dog, they’re better at handling the dog. So the dog’s confidence is already starting at a higher level than if it had somebody that felt like they’ve never held the handle or anything before then. So, you know, we get that, you know, double win really from doing it that way. And we can use the platform for all competing reinforcements that we might have. So other dogs, where you can see my dog behind me. So sometimes we might be in a pet dog. And then exactly the same, we can have the dog walking up between the platforms, and we have the pet dog out to the side. And eventually we can get the dog closer and closer. And then we can be walking across their path. So they learn what it feels like. And then when we’re in the real environment. And there’s a real dog there or real food on the pavement, we can talk them through it first time, and then they start to learn what that feels like what they do in those situations, you know, what works for them and that individual dog as well.
Kayla Fratt 27:23
Yeah, that makes sense. I’ve been, I think I saw a photo on the the Instagram page of skylo seacology, which is another conservation dog trading organization down in Australia, New Zealand, Australia, somewhere in Oceania. I’m pretty sure they’re Australia. And they, they had a photo of they were working with one of their dogs at a farm. And I think there was, you know, there was like llamas and donkeys and stuff. And I think they were just starting to introduce the dog and kind of gauge the dog’s interest around livestock. But ever since seeing that photo, I’ve been kind of obsessed with the idea of trying to take my dogs consciously to farms, and having them search near cattle that are fenced and practicing that as a way. I think a lot of times in this field, and it may be similar in in the past, in the guide dog world, we kind of hope that we’ve just got such a strong reinforcement history for odor and that our dogs are so ball crazy that they’ll just ignore an elk the first time they see it. And I think that’s a pretty big ask for a lot of dogs. And especially, you know, as far as personality goes, the populations of dogs that you and I are working with are pretty different. As far as working dogs go, we’re really looking for dogs that are very attuned to movement and very, you know, ball crazy and pray, or at least enjoying the chase, correlate pretty strongly. So anyway, I’ve been thinking through a lot, you know, how I can be more intelligent in the future as far as introducing my dogs to the concept of, you know, not just searching around people and other dogs, but wildlife and especially, you know, maybe maybe going around some chickens or guinea fowl or something that’s going to really scatter in a very exciting way versus, you know, a couple big Clydesdales. That’s that’s a lot easier even for my my young herding breed.
Tony Harvey 29:11
Yeah, it’s interesting, isn’t it? Well, because I think we’re, we’re very conscious now aren’t we are setting our dogs up for success all the time, and our animals and humans up for success. And sometimes when we work in a more sterile environment, you know, while we’re introducing stuff, it can be quite scary, can’t it to move on to start introducing those distractions, because we don’t want it to go wrong. So we also have to be aware that when we are doing this work, we need to be moving on at a pace which is right for our dogs, but we need to be progressing stuff as we’re going as well and always increasing that criteria of what we’re expecting. And certainly, obviously can go quite fast. When we first introduce, we call it food manner. So which eventually leads up to not eating the food off the pavement. And we’re start with maybe food lined up on a table or something like that. And we just had the dog, you know, a couple of meters away and they would just give food to the dog just for not going towards it. mood and adventure with knock it off. But we’d quite quickly changed that kibble to maybe a crisp or something more real and make it all more realistic just that we’re making sure that we’re constantly increasing that criteria as well as generalized into everything brilliant and saying, you know, this set of rules applies to all of this, in the hope that if something novel comes up, it’s applied to enough things that it’s like, oh, well, this rule applies to this as well. You know, I’ve done it so many times.
Kayla Fratt 30:24
Yeah, yeah, absolutely. I mean, I know like this last summer, I had the privilege of the wind farm I was working on had a lot of free range cattle. And that was a really, really good learning experience for my younger dog. Because we had a ton of practice around cattle, which to him, probably isn’t that different than pronghorn antelope, or elk or whatever. But, and they’re honestly in some ways, they’re just as dangerous if not more dangerous, but at least they’re semi domestic, they’re pretty predictable. And because they traveled in such big herds, I could really see them and kind of pick up the long line and work through managing things before things got out of hand. So I’ve got three questions from Patreon, and then we’re actually gonna gonna be able to wrap up here. So two are pretty closely related. So Taylor asked if there are successful, or specific dog traits that make this trading easier or more successful. And then Megan, from Patreon asked, if you factor the need for Intelligent Disobedience in when selecting a prospect for service dog work. So those are different but related questions.
Tony Harvey 31:33
Yeah, is a really interesting question, especially in our line of work, because we kind of need a variety of dogs with different traits quite often, because we have a variety of clients with different traits. So you know, we might have a person who just doesn’t do so much as somebody else. Or we might have a person who walks really slowly, and they just have a small field that they not, you know, a small field or for their environment that they access on a daily basis. So that could just be the local shops and their local village or town. And then we might have your guide or go into that uses public transport, they fly here, then everywhere they work through different city centers every day. So we actually need a variety of dogs. It’s quite well known. I think the breed we use most of the time are your kind of Labrador, Labrador Retrievers, or your golden retrievers, and across those both. In the UK, we sometimes use German Shepherds as well. Labradoodles. And we sometimes use some pure poodles as well. But generally, the dogs that we use most are your Labradors, and golden retriever crosses, historically, I guess, because they’re a lot easier to work with. It helps that the lab, you know, brings out that food in us and our dogs as well. And that really helps with the kind of training we’re trying to do now. And in terms of guests, you know, the obvious thing that we have to remember is that we need the right size for our dogs as well. So we can’t use Jack Russells and things. They’re just too small, it doesn’t work. I think there’s an Australian advert where some guy does have a Jack Russell as his guide, but ya know, size is quite important needs to be a certain size and the size of somebody. So all of those things, but yeah, we’re looking for temperament. And in the past, I know there’s schools around the world that have used different dogs that Dalmatians have been used before boxes have been used before. There’s another large breed that I can’t remember, but it was fairly, relatively short lived as well, about seven years or so. So when we’re breed in just a little bit about the breed, and we just the main thing we’re looking for is health and temperament, really, so we’re really built breed into make sure we got our healthy dogs, and the right sort of temperament, which is we don’t really want the chase in there. You know, there’s lots of those other things, we don’t want to that high scent work that you’re probably really looking for, we don’t really want that either, we just want them to be quite into the handless or get quite handler focused so that we can really work on that train and progress it in terms of when we start to introduce the decision trees and all of that Intelligent Disobedience. That would be we’ve got about 37 behaviors, we call it step. So it’s standardized training for excellent partnerships in our, in the UK guide dogs. And once our dogs has kind of got that behavior, we will probably start to want to introduce all of those other aspects really quickly. So we start to introduce, okay, so you know what to do in this situation. But then if it’s slightly different, and this happens, then you do this or and then if this happens, and you do this, so we start to introduce it really early on. So with just trying to think what a good example might be, because we do location objectives as well. And what we do is we teach them to find specific points, and we can train the guy to go in is then to make a new specific point for them. And we sometimes call it magnetizing an object so it could be like a postbox or a shop door. And at first this means nothing to the dog. And then we train the dog that it’s something really important to you and we you know we back chain it’s they learn to find it, learn to find it then to find it. But then quite quickly, we only want to find it when we give them the cue for that particular object. So that might be the door or whatever shop it is hesitant to shop names because I know they’re going to be different in all the different countries He’s. So you might say Tesco, you might have heard of Tesco as a Walmart or wherever it is. So you might say Walmart. But there’d be days when you just want them to walk past it. So as soon as you taught them the cue, and they know how to find it, in the absence of that cue, or with another cue, you want them to walk straight past that into the next objective or the next obstacle or curb. And we could teach that straight on means you keep going past that. And the way we might do that in the training arena to begin with, is we might have a crossing box and you’re left with a platform there or something, we might have a chair on your right, and then we might have just a platform straight ahead. So we could teach, the crossing bots could just be box, and then they go off to the left, the chair could be chair, find it on the right, or we could say straight on or nothing at all. And that just means keep going straight until you get to that platform. So even then, when we’re teaching stuff like that, we can still teach it in that safe environment, you were talking about having your fenced animals in the world. And that would be very much the same for us, we have a safe environment. This is what it means in this context. And then when we get onto the streets, it’s all exactly the same. We just moved all the while we’ve hired, we’ve raised the criteria, but we’ve changed the environment, really, but it still means the same. Does that answer?
Kayla Fratt 36:09
Okay? I think so. I think so. So yeah, it’s not so much that we’re, as you’re selecting a dog, or you’re looking at like a nine month old puppy, you’re doing pests to see how you know, you’re you’re just working with whatever they’ve got to teach them the skills they need for the partnerships we have.
Tony Harvey 36:28
In the UK, we have our own breeding programs, so we don’t select dogs, as it were, we do the breeding, and then the pups come through and we assess them continually coming through. So at any stage on the guide dog journey, when they’re a pup, when they’re in training, or when they go on class where the guys are gonna, they can potentially be repurposed to another job. And we have buddy dogs in the UK, or they can be withdrawn in and become a pet dog as well. So they’re always being assessed. But the idea is that every dog that we bred, will hopefully be a guide dog. So that’s kind of how we view them. And we work to train as many as we can to be the guide dogs.
Kayla Fratt 37:02
Okay, yeah, that makes a lot of sense. I think. You know, it is interesting, I think, because so many of us in the conservation dog world, we’re working with rescue dogs, or, you know, there’s not. Not yet is there anyone who’s breeding dogs specifically for conservation and detection, dog work, and there’s so many different, you know, similar to what you said, as far as the role can look really different from John for dog, each dog. You know, I can see when I’ve talked to some handlers or some handlers who really prefer Spaniels, they like dogs that are highly, highly independent, and very much so are going to go out and do the search on their own. Versus I work with Border Collies, it doesn’t get a lot more codependent than a border collie. And we work really, really closely together. And they take a lot of direction really, really nicely. And I think that’s probably more important and does factor into this. Like, I would imagine, I have to do a little bit more work teaching my dogs to ignore my body movements that someone who’s handling a spaniel may have to because of the breed that we’ve each selected. But it seems a little bit more like a personal preference thing. And then just working with the dog you’ve got versus something, I wouldn’t necessarily rule a dog out. Because they struggled with.
Tony Harvey 38:20
I guess that’s one of the big parts for us as well. It’s our matching process. And when we match our guide dogs to our clients, that’s probably the biggest part of the job. If we get that right makes everything else easier. And I mentioned labs earlier being foodie, we know they’re not all particularly foodie. And there’ll be some dogs, which probably need that higher level of reinforcement when they’re out in the real world working. And there’ll be others probably don’t need hardly any food at all. And we have clients that will be more than happy to use food every day on every curve and every objectives. But when we’ve got those clients as well, which might not be comfortable, for whatever reason, potentially, they’re wearing suits to go to work, and it’s a bit greasy on their hands. It doesn’t really matter. But we can match around that. And so yeah, that’s why we need all those different dogs with all those different attributes. And as we’re going through training as well, we can train the clients those needs. So when we’re getting towards the end of training, we do try and reduce a lot of the reinforcement food that we’re using with our dogs. And then when they transition over to our visually impaired client, again, we raised all that food level right back up again, just to build up the Trust Bank and get them working together. And then we would teach them how to reduce that level of food again, that level of reinforcement.
Kayla Fratt 39:26
Yeah, that makes sense. And then our last question from Jana, I think actually relates to exactly what we were just talking about, as far as four different types of dog personalities. How do does he do you, Tony, approach to this sort of training? And do you have any problem solving examples? So maybe, like, we could take two different extremes, if there’s anything you can think of as far as a dog who maybe is much more much, much more foodie versus much, much more handler centric. Are there differences in how you train this
Tony Harvey 40:00
Um, well, yes, yes and no. So essentially, we’d have the same training program, but how you would work with that individual dog will be different. And I can think about situations actually, when I was in New Zealand, and I had probably that exact scenario. So I had two dogs, one who, you know, you could train, train, train train training, you had to stop the session, and the other who’d be like, one or two days, and then that’s it for me, I’m gonna, I’m gonna lay down now, that’s all I want to do. So I was doing locating the crossing boxes for this and teaching them to target these objects. And so yeah, with one dog, I could do 1020 reps. And that’d be absolutely fine. always wary not to just keep going to always want to end on the successful one. I always say as well, it’s like, it’s like when you’re reading a book, isn’t it? Because we always say, end and successful bit or good bit. And I always think if you’re reading a book, and it’s really good, you can’t wait to pick it up again. But if you read in a book, and by the time you stop, you’re really bored, you just don’t pick it up again. And that’s how I try and think about the train. And so with the other dog, I would try and do just a couple of reps, think that was good. And then I’d stop the session there. And then I’d give them a break bit of a play. And then I’d go back and just do a couple more reps. But generally find with those types of dogs, once they kind of learn the process and the kind of way you’re teaching, you can normally start to get more and more out of them as the training progresses. And it’s not, it’s probably obvious, but it’s not always about the food for them at all. And maybe it’s about, you know, just working with a person or working with. So having that whole time and bonding and listening to your dog as well responding to their needs and their wants is probably a big part of getting them to work with you and getting that buy in. And then progressing from there and building up as much as you can. So same training program, different approaches, I guess you just respond to them in different way on those levels of schedules of reinforcement and everything like that.
Kayla Fratt 41:43
Yeah, yeah, that makes perfect sense. Is there anything you wanted to circle back to or expand upon a little bit more anything? I didn’t ask you about this. You wanted to make sure we covered?
Tony Harvey 41:54
Yeah, no, I haven’t a Nabi on something. So I’ve talked about tactics a lot more than I have on this particular one. But that’ll be more of the client side of things. And the only reason I talked about it is because it ties in so closely with the the animal training we do now. And I think I just think sometimes it’s really easy to forget the humans in all of the thing that we’re doing, isn’t it and we concentrate so hard on getting it right and doing all of that. Sometimes a human comes in, I don’t know whether you have to hand over dogs that much or teach other people. So it’s always bearing in mind that when we are, it’s all about that. It’s exactly the same as it’s breaking down the behaviors, reinforcing the bay was we want making sure we’ve given them everything we can to set them up for success. So no, I think I’ve mentioned that certain puppy needs to go on about it, too. Yeah.
Kayla Fratt 42:38
No, I don’t know if there’s too much. I think it’s just it’s really interesting, kind of think through how I’m going to take what we’ve talked about and set up some set up some good training sessions, you know, and thinking beyond just adding distractor sense, which is something we do really, really pretty early on for our dogs, like with my dog with my puppy. niffler I think is very, like his third session where I was teaching him to find dead bats. We I started having the same treats, he was being rewarded with intense. So he was having to ignore that. But I think and I think most of us in the sent work world do that. And now I’m I’m really excited about this conversation just kind of thinking through. Okay, so what are some of the other ways that I can layer this in and teaching, you know, part of it is proofing and really just being like, okay, when you’re searching, you’re searching No matter what other competing reinforcers are around? But also then, yeah, like, what is my what is my emergency recall queue going to look like? And how do I start teaching them that this one, even if you’re in older you’re going to listen to and then most of you know the rest of the time? So yeah, I think yeah, that decision tree framework. Maybe I need to sit down and write out what that decision tree would look like. If I was presenting it to my dog, and then and then from there, I think you’ve made it pretty clear how you would actually teach that it’s been illustrated quite nicely.
Tony Harvey 44:02
I think that’s a good point, is it better planning up our side sometimes would really help. And we talked about distractions, and something that I’ve found really helps me because if we think about a distraction, and your dog goes towards a distraction, our natural feeling is they’re doing something wrong, or they’re being disobedient. But if we actually frame it for what it is, which is that competing reinforcer, they’re not doing anything wrong. They’re just going into something that’s more reinforcing than what we’re offering before. And when we start to think of it in those terms, it becomes a lot easier to think about what plan we have to put in place to work through that. So you know, this dog finds his other dog really reinforcing. So I need something as or more reinforcing to get what I want out of it. And I think that’s the same in all the worlds that we work in.
Kayla Fratt 44:43
Yeah, yeah, absolutely. Well, Tony, thank you so much. Where can people if they’re interested in learning more about the work you do or following you online, Is there anywhere that they can check out?
Tony Harvey 44:55
In terms of Guide Dogs it’s just the guide dogs UK website, which is easy enough to find through Google I, myself and my wife have a Facebook page. It’s called click Fit training where you can see some of the goats and some of tui behind me on there. And some of the training that we’ve done, I think there is a, there was a little film on there of Emily, my wife training a goldfish to score a goal as well from a few years ago. So yeah, that’s the main pieces from a really tagteach stuff, obviously, this tech teach websites online as well, which is always really good to look at and start seeing some of the videos that they’ve done.
Kayla Fratt 45:27
Yeah, well, great. And yeah, and again, thank you so much for coming on and chatting. And to our listeners, thank you all for listening. I hope that you learned a lot and are feeling inspired to get outside and be a canine conservationist in whatever way suits your passions and your skill set. You can find show notes and donate to Canine conservationists and join our Patreon so you can ask questions of guests like Tony over at Canine conservationists.org We’ll be sure to have links to all of Tony’s social media and the things that he mentioned will link that paper with the pigeons and the lights as well. So you can find all that there until next time.
Kayla Fratt 46:23
Are you on Patreon yet, if you love this podcast and want to support it in the long term, Patreon is the way to go. I spend hours per episode researching guests writing out questions, recording interviews, posting on Patreon to engage with our patrons about all of those, cleaning up the audio and putting together all of the promotional materials. Even with the help of volunteers. This is an enormous task that takes up a ton of my time. And right now I’m not paid for it for just $3 a month, you can support this show while also getting access to our exclusive detection dog training video help calls, which happened once a month, our learning club calls which are currently quarterly but I’m hoping to move to monthly and a lot more, you can join the fun over at patreon.com/canine conservationists or using the link in our show notes. You also may want to share this with anyone else you know who is interested in getting involved in the field of conservation detection dogs, because this is hands down the lowest cost way to get as much mentoring and assistance and joining the community of other professional and aspiring conservation detection dog handlers. And you’re gonna really enjoy it. See you there.
Transcribed by https://otter.ai