Guest: Kiki Yablon
Kiki Yablon is a Karen Pryor Academy Certified Training Partner, Certified Professional Dog Trainer-Knowledge Assessed (CPDT-KA), Karen Pryor Academy faculty member, and co-instructor for Susan Friedman’s BehaviorWorks, and holds a master’s degree in applied behavioral science.
Kiki came to clicker training in 2005 as the novice owner of an under-socialized adolescent shelter dog, Pigeon. She had the good fortune to live next door to a marine mammal trainer who introduced her to Karen Pryor’s work, which in turn led Kiki to her first mentor, Laura Monaco Torelli. After briefly wandering in the unregulated training wilderness, Kiki attended and graduated from KPA. She then worked as Laura’s righthand woman for eight years before focusing on her own business, Kiki Yablon Dog Training. Some of Kiki’s training interests include excessive barking, proactive puppy raising, loose-leash walking, unusual problem behaviors, and “lazy” solutions that let a well-designed environment do most of the work.
In a previous career, Kiki was a magazine editor. She currently puts those skills to use writing about the application of behavior science to dog training on her well-trafficked blog. Kiki lives with her husband and their dog, Finn, in Chicago, Illinois.
In this episode we discuss:
- What is stimulus-stimulus pairing and how it works?
- What is a “classically conditioned recall” and why might we consider calling it something different?
- Revisiting the differences and interplay between operant and classical conditioning.
- The difference between describing a procedure and describing a process.
- Lots and lots of examples of how to apply this concept in different training applications.
Links mentioned:
- Kiki Yablon Dog Training website
- Kiki’s Instagram
- Episode #103: Dog Emotions in Practical Training with Kiki Yablon
- Episode #144: Pet Trainer Roundtable
- BehaviorWorks LLA course
Support The Show On Patreon!
Thank YOU for being a part of our community! I couldn’t do what I do without you. To become a tangible supporter of the show, please visit our Patreon page and share the word in your circles.
Episode Transcript
Kiki Yablon: [00:00:00] You know, my dog Finn is laying on the bed and thumps his tail at me and I pet him. And while I’m petting him, the thumping stops. And then when I stop petting him, he thumps his tail. And then I pet him and then he more quickly thumps his tail each time.
Hannah Branigan: And if you don’t respond, he puts his nose under your elbow and flips your arm up so your hand happens to land on him.
Kiki Yablon: Right. If you extinguish tail thumping, you get something else.
Hannah Branigan: Yeah, that’s the stimulus response.
[intro]
Hey there, fellow training nerds! You’re listening to Drinking from the Toilet. If you like to geek out about combining the science of behavior with positive reinforcement philosophy in real life, you’ve come to the right place. And I’m your host, Hannah Branigan: [00:01:00] teacher, trainer, podcaster, and author of the book Awesome Obedience and its companion Awesome Obedience: the Field Guide, which are both available from clickertraining.com.
So this week we are hanging out with returning guest, Kiki Yablon and we’re talking about a strategy or collection of training strategies that fall under the umbrella of stimulus/stimulus pairing. So don’t let that term intimidate you, because while of course this conversation is pretty nerdy tying into what we know about the science, we’re also really all about practical application here. And if I recall correctly, we go into a lot of real life examples of the concepts. So I really think you’re going to get a lot out of it no matter what area of dog training you fall into.
But before we get into that, I gotta send a shout out to some awesome folks who support this podcast on Patreon. So I want to personally thank Michael R, Jules C, Serena E, Gabby H, and Jonathan P. You’re amazing and I love you. Our next live patrons-only [00:02:00] Q&A session will be on June 8th. So if you wanna join us, you can ask any questions that you have, get help troubleshooting your own training with your own dogs, and even see a technique or exercise demonstrated as long as it’s something in my repertoire. If you’d like to support the podcast, get your questions answered, get access to our super-secret extra podcast episodes, you can go to patreon.com/DFTT.
So Kiki, who you will remember from multiple other episodes, she’s a Karen Pryor Academy Certified Training Partner. She is a Certified Professional Dog Trainer – Knowledge Assessed. She is a Karen Pryor Academy faculty member, a co-instructor for Dr. Susan Friedman’s Behaviorworks and holds a master’s degree in Applied Behavioral Science. Which all adds up to solid credentials with that sweet combination of academic and scientific background with practical application. She knows her stuff and she’s always at the top of my short list of folks when I need somebody to [00:03:00] check my work from a training perspective.
So I won’t give you too many details, because we talk all about it in the conversation. But a large part of our conversation centers around this interplay between operant conditioning and classical conditioning and how we can leverage one or the other with our training and really kind of take some shortcuts – but not the bad kind of shortcuts, not the like, “Oh, you’re cutting corners now to make a quick buck” kind of shortcuts. Clearly we don’t do that kind of thing. It’s more like, “Hey, here are some naturally occurring phenomenon that we can harness to make our training simpler and so much faster and smoother.” Which, as a lazy dog trainer, is what I am all about.
And I really think that that’s what is at the core of stimulus/stimulus pairing, for our purposes anyways.
So enough about that. I’m gonna stop talking and I’m gonna pick back up with a conversation in just a second. [00:04:00]
[episode begins]
Hannah Branigan: I think a lot of folks are familiar with the technique that’s often described as “classically conditioning a recall.” I feel like I should put air quotes around that, which maybe will become more clear later as we dig into this a little bit.
Can you say, say something about that? Like describe what’s going on there. Like how’s it playing out?
Kiki Yablon: Okay. So usually I think it is some version of “make the cue happen, whether that’s saying a word or blowing a whistle, and then make the food happen.” So just cue/food, cue/food, cue/food or whatever the reinforcer might be. Cue/whip out a tug toy. Cue/run like Phoebe on Friends. Whatever reinforcer you’re using to reinforce this important [00:05:00] behavior. And that’s often referred to as classically conditioning a recall.
And I wanna say that there’s some debate about whether that’s the right thing to call it. I don’t think it’s the right thing to call it, but there are decent arguments on the other side. Basically, there’s a school of thought that– well, let me back up–
Hannah Branigan: Actually, let me repeat back. When I’ve done this and taught this – and I just wanna make sure we’re talking about the same thing – what I usually do is I’m gonna stand in my kitchen. I’m gonna have a can of cat food and a spoon, and I’m gonna say, “Brownie, come,” and then I’m gonna just spoon cat food into Brownie’s face and he’s starting and finishing right in front of me. So it doesn’t seem like there’s any kind of a recall there at all because he’s not [00:06:00] coming from anywhere. I’m just saying “come” and then stuffing his face with good food. And so that’s why I have in the past called it “classically conditioning a recall” because I had no criteria other than like letting me stuff your face with food. He’s not doing anything–
Kiki Yablon: The contingency is between one stimulus and another stimulus. If this happens, that’s going to happen.
Hannah Branigan: So he’s gonna get food in his face no matter what he does between when I say come, as long as– I mean, I guess if he were to leave the house, I wouldn’t be able to to do that practically, but, but I’m just gonna be handing out food to anyone in the vicinity once I’ve said the cue.
Kiki Yablon: Yeah. The way I often explain it to clients is that you’re just a robot that says “come” and then drops a bunch of treats or whatever. [00:07:00] And the dog’s gonna go, “I see. It seems to me that when you say that word, then treats fall out of your hand. So I’m gonna use the behavior I have to be there when it happens.”
Hannah Branigan: Yeah! “I wanna get to the drive-thru window in time for that food to be available.” Yeah. Okay. So I’ve called that classically conditioning because I tend to use that phrase to describe a technique or strategy that I’m using where I’m really not paying attention to the dog’s behavior in any way. I’m just making a noise or doing a thing and then, like you said, making food. Hah! That sounds like I’m regurgitating it at them!
Kiki Yablon: Blegh!
Hannah Branigan: “And then I vomited into my dog’s mouth!”
Kiki Yablon: Right. I mean, I’ve called it that too. I’ve called it that too. [00:08:00] I don’t always call it that anymore. And we’re gonna talk about why, but yeah.
Hannah Branigan: So now circle back. Where were you going with that?
Kiki Yablon: So then what we usually start to do is when we start to look for some response when we say the word. And there’s a little gap in time between when we say the word and when we reach in our pocket or when we start to regurgitate the food. Right?
Hannah Branigan: I mean, we do train in a natural way. So just like a mama wolf? Yeah. That is how I live with my dogs.
Kiki Yablon: Yes, of course.
So then there in that little gap, we start to look for something that we call “anticipation” or “happy face” or whatever that indicates that the dog has made the connection that the sound means the food is going to follow. And then what we might start to do [00:09:00] is give our cue when the dog is not looking at us or has their back turned to us and see if they turn around, and then off to the races with increasing distance and distraction and all of that stuff.
So there’s a school of thought that that is classical conditioning because the procedure is to pair one stimulus with another stimulus.
And the confusion about this sort of comes from the fact that the terms classical conditioning and operant conditioning describe two things. They describe the procedure that we do, like what the trainer does or what “nature” does, what happens in the environment, and they also describe what the animal learns. They describe the learning process. So we could say procedures and processes, [00:10:00] right? Learning processes and procedures are things that we do typically. But we could also say the environment is doing them. Thunder proceeds lightning. So God is doing classical conditioning.
Hannah Branigan: I mean, that does– It feels like hair-splitting at first, but the longer I spend thinking about these things alone in my house, the more I recognize that it really does matter. Like we’d say, “Oh, it doesn’t matter what you intend to reinforce; it’s what your dog find reinforcing.” Like there’s lots of places where there is a disconnection between the trainer’s intention and what the learner is actually experiencing. And it seems to me this fits into that as well. Like there’s the technique I’m applying, the procedure, which is what I’m thinking about, what’s going on in my head, which may or may not be reflected in my learner’s behavior. Hopefully it is! Like my goal is for those two things to come as close together [00:11:00] as possible. But they are different.
Kiki Yablon: Yeah. And actually the terms reinforcement and punishment have the same problem in that they describe a procedure and a naturally occurring process. So it’s not incorrect to say, “I reinforced the sit by giving the dog a treat” if you’re talking about procedures, but if you’re talking about processes, to know if you reinforced the sit, we need to measure the frequency of sit over time when the treat is the consequence and see if it increases, usually in rate, like how often when you say sit does the dog do the sit? Or how often when you’re standing in front of the dog with a treat pouch on does the dog sit? Or whatever the antecedent is at that time.
So you’ll hear people correcting each other like, “Oh, you can’t say that’s reinforcement because you don’t [00:12:00] know what happens next.” And in some– So if you’re talking to someone about what procedure to do, it is useful to say, “Reinforce now.” Right? But if you are talking about whether your procedure worked, then it is not useful to talk about the procedure. You wanna measure behavior over time and that’s when you’re gonna say “it was reinforcement.” So I think calling things things is useful in some circumstances and not useful in others.
Hannah Branigan: And I mean, do you ever feel like it’s just a moving target and every time you think you’ve got your head wrapped around it, you find out that you don’t and like your words are wrong?
Kiki Yablon: No. I mean, I think we can hold multiple things in our heads at once, but I also think that in terms of like being in the business of educating newer trainers and [00:13:00] things like that, or teaching clients how to get the behavior that they need or want from their dog, what’s more useful than any of this stuff is just saying what you do.
Hannah Branigan: Yeah. So that’s where you would just say, “I delivered food. I gave the dog a spoon full of cat food” or whatever, versus–
Kiki Yablon: Yeah, “I said the word and then I gave a spoon full of cat food.” And I think that that would help in a lot of ways. It would un-obscure what is actually being done. Because I’ve seen– you see client reports come through, say that they’ve done counterconditioning and desensitization, right? I have one idea of what that looks like. Somebody else might have a different idea of what that looks like. So what I wanna know is, where was the dog? What was the trigger? When it appeared, what did you do? Right? Like, those are the things that are actually useful. And then, as a [00:14:00] professional, yeah, I think you wanna understand what’s making that procedure effective or not.
As you know, I have a relatively new degree in applied behavioral science and there’s this seminal paper called – I forgot what it’s called – It’s like Some Dimensions of Applied Behavior analysis [Baer, Wolf and Risley, 1968]. Basically this paper from the late 1960s outlining like what is Applied Behavior Analysis, like what makes it Applied Behavior Analysis.
And two of those seven things are that we use technological description, which basically means that someone with the right background should be able to read your description of what you did and do it. So if you just say that you did counterconditioning, that doesn’t give me enough information to do what you did. If you say that you paired this stimulus with [00:15:00] this stimulus with this timing at this distance from another dog, that tells me what to do.
And then another one of those dimensions is what’s called being conceptually systematic, which means you can tie it to basic principles. You can say, “This is why we would do this; this is why we would think this would work,” more or less. And I think that those would be good values for dog trainers to adopt for practical reasons.
Hannah Branigan: Just from a communication standpoint, I think about– You know, someone will reach out to me and they’ll say, “Well, I tried that technique and it didn’t work.” And that doesn’t actually usually tell me a lot. Like it could be totally valid and then I need to know why and we have to ask more questions to kinda get to the bottom of that. But what I really wanna know is like, “Well, what did that look like?” [00:16:00] I mean, I need to know that either way. I need to know what did that look like? Because the when, what was the context? What did you do? When did you do that? What did the dog do? What did that look like? I mean, even if they said “we did it and it did work,” like I still need to know the same things if I’m gonna repeat it or if I’m gonna troubleshoot it or otherwise figure it out. Because the timing matters. The specifics matter a lot and we can’t really get around that.
Kiki Yablon: Right. Like “I stood right in front of the other dog and I offered a Zuke five seconds after the other dog barked.” Yeah, no, then we can understand why that didn’t work.
Hannah Branigan: Yeah. When you say like counterconditioning, like the first thing that pops into my mind was, “Well, I’m hoping to countercondition.” That’s usually like– when I walk in and say, “Well, I’d love for this to become counterconditioned, but that may or may not happen.”
And then yeah, someone will say, “Oh, well, I tried counterconditioning and it didn’t work.” And then when they describe it, “Well, I went to the dog park [00:17:00] and I stood in the middle of the dog park while other dogs came up and sniffed my dog’s butt and he wouldn’t take food, so it didn’t work.” And I was like, “Oh, yeah, imagine that. Okay, maybe we could do something. Let’s refine that a little bit. Let’s see if we can tease that apart.”
Kiki Yablon: Yeah, I like getting that information cuz then I’m like, “Okay, we haven’t done what I would do yet.”
Hannah Branigan: That actually does feel better than, “Oh, you did exactly what I would do and this still didn’t help. Yeah. I’m out. Uh, lemme refer you to somebody.”
Kiki Yablon: Yeah.
Hannah Branigan: So that’s not off topic, but let’s steer back towards what we were talking about.
Kiki Yablon: Yeah. Well, should we talk about how we started talking about this, which is that I gave a presentation at Clicker Expo Live back in January, which you were kind enough to sit through twice, I think. Once to help me sort it out and once when I actually gave it. And what the presentation was about was sort of [00:18:00] client-friendly shortcuts to teaching what are operant behaviors like recall and dropping things and some of the ones that people often say they’re classically conditioning, but also lots of other operant behaviors, by doing what I think is most useful to call stimulus/stimulus pairing. So pairing one event/object/location with another one in such a way that the behavior that we want emerges in between them in the gap.
Hannah Branigan: So one of the things that I took away from the Living and Learning with Animals [LLA] course, which is– It’s one of those things, like so many things, where you learn it and you’re like, “Well, yeah, that’s really obvious. I already knew that,” but you didn’t necessarily have it in words. It was that when we’re talking about the ABC, antecedent behavior consequence, [00:19:00] our three term contingency, part of how you can identify it is that the A and the C are things that happen in the environment. So those are stimuli. And then the B is whatever your animal does.
And so then in my mind, I’ve got this little like– Actually, it reminds me a lot of how Emelie and Eva say like “This is your dance space. This is my dance space.” And so my dance space is the A and the C. And so those are stimuli, those are things that are happening outside of my dog. And then the B is the thing that my dog is doing. And so we’re just kind of nipping off the B here and talking about the A and the C kind of right next to each other.
Kiki Yablon: Yes. But with an eye to the B. So that’s where I think the classical conditioning way of talking about doing these things is not as useful as it could be.
So we should back up and say, another thing you would learn in LLA [00:20:00] is that operant and classical conditioning are always happening at exactly the same time and that the distinction between them is kind of a textbook man-made thing. It’s academic to a certain extent. And within behavior science, there’s a whole range of thinking.
I mean, the mainstream is still what you’re gonna learn and you should learn like those basic ways of thinking about things before you go wandering off into this other stuff if you’re new. But there’s a quote that I put in my presentation from a guy named John Donahoe who – not to get too far into it, but has basically proposed like that there’s what’s called a unified theory of reinforcement that like operant and respondent learning occurs kind of by the same mechanism.
But basically this quote kind of stood out at me from a presentation that I [00:21:00] watched by him. He says something similar in writing. I’m gonna paraphrase it. Pavlov’s dog, right. Strapped into a harness and then – I guess he didn’t use a bell, but he used it a metronome. So Pavlov’s dog might prick up its ears when it hears the metronome tick and that is followed by the reinforcing event, which would be food. Did the food occur because the metronome made a sound or does it occur because the dog pricked up its ears, which is being discussed as an operant behavior here?
What he’s saying is that the first time that this happens, there’s no way that the dog would be able to distinguish whether the food was happening because the metronome sounded or because it pricked up its ears. But that moment when the food appears, the classical and operant conditioning are the same [00:22:00] procedure. But then the differences in those procedures emerge over time.
So when we say “come” and we produce food, we are going to elicit food-related reflexive behavior like drooling, things like that. And then we are also going to probably evoke behavior that saying something in front of a dog while you’re holding food already evokes. If you make a noise, you’re often gonna get sort of an orientation response, especially if you’re close to the dog and there’s not like a lot of distraction.
Hannah Branigan: Yeah. It’s a high-contrast stimulus. Dropping the tray in the cafeteria when nobody else is talking.
Kiki Yablon: Yeah. [00:23:00] Right. If you’re in a quiet room and I say something, you’re gonna probably turn around and look at me, which I would say is an operant behavior. There’s some discussion about that too. But once that’s a behavior that once it hits the environment, it’s definitely operant. Like you might turn initially every time you hear any sound, but you learn which sounds to turn to and which sounds not to turn to or you would be going crazy. You can also learn to not do it. I’m gonna just pretend I didn’t hear that person call my name because they are unpleasant.
So basically with the food, you’re gonna capture that orienting response or that looking at you response. And you’re gonna capture it in a particular location, which is in front of you. And then you’re kind of off to the [00:24:00] races. So you say the word, you deliver the food and there’s gonna be necessarily a gap in time there.
And I don’t know, this is just wild, this is something that just occurred to me now, but maybe that’s why the types of conditioning– you know, with classical conditioning, you’re supposed to have one proceed the other the onset of one is supposed to start to happen before the onset of the other. That’s the most effective way to do classical conditioning. Maybe that’s why.
But you’re gonna reinforce orienting and you’re gonna elicit things related to food like drooling that aren’t usually subject to changing with consequences. And then if you are setting up your training session the way that a lot of us are trying to do more lately, which is that you’re not raising your criteria [00:25:00] before the dog is ready so that they can learn, “oh, that didn’t work.” You’re not setting them up to fail. You’re trying to set them up to do what we call errorless learning. Not meaning that there’s no errors, but just meaning like mostly we’re setting them up, we’re setting the the environment up and the antecedents up so that when they do the behavior, they’re gonna do the behavior and they’re gonna get the reinforcer. Then it’s not going to look very different from a classical conditioning procedure because the antecedent occurs = you’re gonna give the reinforcer.
Hannah Branigan: No, that makes sense. And so one of the ways that I have thought about this – I don’t know, I found it helpful. You tell me if I’m taking a wrong track – but like when you said you capture the ear prick along with everything else, that used to get me tangled up kind of easily and then I [00:26:00] started thinking about it as, “Well, it’s really the system is capturing all of those things at the same time. And the system doesn’t have a way of knowing what of all the things that are happening at the same time are the relevant ones. Like what’s significant here.” Right? But then presumably over time, you sample a bunch of these moments over time and then your system is overlaying those and some of the things are gonna be consistent from picture to picture to picture and others won’t be. And so you’re eliminating the things that don’t matter. It’s like, “Oh, well this time, I happen to have tilted my head to the left. Oh, this time, a bird made a sound outside. Oh, this time–” And if it only happens inconsistently in that set up, it probably isn’t relevant.
And of course none of this is happening on a conscious level for any of [00:27:00] us. It’s just capturing–
Kiki Yablon: I mean, it can happen on a conscious level, but it also doesn’t require awareness. That’s a common misunderstanding, is that operant behavior is often called voluntary behavior. But it’s often not voluntary. And that doesn’t mean it’s not operant. I mean, it’s not fully voluntary, as in it’s not conscious. People in comas have been operantly conditioned to move more by playing music that they like. Awareness is not a requirement. Like we learn stuff through operant conditioning all day long. I can’t tell you how many– Did I voluntarily just like shift my weight so that my back didn’t hurt anymore.
Hannah Branigan: It’s way deeper than– I mean, one thing that I often think of is if learning had to be conscious, then marketing would not work. But marketing’s really powerful!
Kiki Yablon: Right. That’s a great point!
Hannah Branigan: Our [00:28:00] behavior can be influenced well without any of our awareness or understanding of it. We just do things and we don’t even know. So much social conditioning works that way.
Kiki Yablon: Yeah! So, just to follow out what I was saying a little bit, if what you’re interested in is an operant behavior, then to me it makes sense to talk about the processes that are happening that would influence operant behavior that are happening while you’re doing this procedure.
The other thing I was gonna say is, when we were talking about ears pricking along with other stuff, I recently had an interesting discussion with some colleagues about this, like tail wagging. Somebody referred to tail wagging as respondent. And I said, I don’t think it is. So I gave an example. I made a little video where my dog Finn [00:29:00] is laying on the bed and thumps his tail at me and I pet him. And while I’m petting him, the thumping stops. And then when I stop petting him, he thumps his tail. And then I pet him. And then he more quickly thumps his tail each time.
Hannah Branigan: And if you don’t respond right, he puts his nose under your elbow and flips your arm up so your hand happens to land on him.
Kiki Yablon: Right, if you extinguish tail thumping, you get something else!
Hannah Branigan: That’s the stimulus response between the two–
Kiki Yablon: And so we often might say like “he’s wagging his tail because, because he’s happy.” Well, why is he happy? And I don’t even know if he’s wagging it because he’s happy. Wouldn’t he wag it while I’m petting him instead of– which sometimes happens as well.
But the other day then I noticed that I was making breakfast [00:30:00] and I set my breakfast on the counter and he wagged his tail at me and then I of course had to give him a piece of scrambled egg. But yeah, there’s always a bunch of different behaviors happening at once. They might not be our target behavior, but they come along and I think all of us have had a moment where we noticed that we were getting something besides our target behavior, such as a bark in the middle of your down.
Hannah Branigan: Oh, so you’ve been watching my training?
Kiki Yablon: [laughter] No, just my own training. A bark while the dog is laying down. And that comes because you set the environment up so that the dog barked the first time by putting something in the environment that the dog has a reinforcement history for barking about or whatever, and then [00:31:00] it got captured along with at the same time as the down. But we could say the same thing about behaviors that you want.
And I think there’s different trainers– You’ve talked about this. I think I first was made aware of this from your Obedience FUNdamentals DVDs that you’re like “capturing the emotion along with the behavior,” an older way of talking about it. Or what is Sarah Owings calling it? Like “emotional fluency” or something like that. But all of that is basically those other behaviors are consistently happening at the same time and getting reinforced.
Hannah Branigan: The “ears up, eyes bright, mouth open,” behaviors that we call happy or enthusiastic happen at the same time as the heeling or the sit and get reinforced part of that picture.
Kiki Yablon: And to say that is not to say that we are training dog actors [00:32:00] who don’t really feel the way that we think that they feel. It’s just that emotions are responses. They are a bunch of things that our body does and some of those things are evoked by certain stimuli automatically, some like heart rate and things like that theoretically, and then some of them are reinforceable. Some of them will in increase if the outcome keeps being good.
Hannah Branigan: So we’re talking about setting up those events in the environment, our dance space, the parts we have some control over so that the behavior we want to happen happens in between.
Kiki Yablon: Including the behavior that we call “happy.”
Hannah Branigan: [00:33:00] If that’s what we want, yeah.
So here’s another example that maybe fits here. This is one I was just working on last night. So I’m teaching the pony to stand on a mat and I go about this a lot of different ways, but one of the ways that I was doing last night was I had two bowls set up on either end of the area we’re working in and his big mat in the middle.
And I started this with an enormous mat. It’s like four feet by four feet. It’s like half a piece of plywood, but not plywood. And I put some treats in one bowl and then I went across and I put treats in the other bowl. And in order to walk after he ate the food that was in the first bowl, he had to walk across the mat to get to the treats in the second bowl. And the mat was big enough that he couldn’t– I mean, he could have gone around but why would he? It would’ve been effortful. And this is a pony that is nothing if not efficient. So I did that and I put [00:34:00] treats in the first bowl, walked across, put treats in second bowl, walked across. And so he’s walking and eating these treats and he is walking across the mat, and he’s walking across the mat, and he’s walking across the mat.
And then I put treats in the first bowl and he ate those treats and I walked partway and paused and he continued to walk. When he stepped on the mat, because it was in his path, I marked and I put food in the second bowl.
And then we did the same thing back the other direction. I paused and he kept walking and stepped on the mat. I marked and I went and put food in the second bowl.
Then I started pausing earlier and earlier and earlier, until he’s walking ahead of me and stepping on the mat.
And then – and this is the piece that I always look for in this moment – he walked almost all the way to the mat, paused and like looked at me. And then stepped onto the mat and looked at me, like the significant “AHEM!” look, which is very different from a – I used to think it was like a reflexive look, but it’s the look of [00:35:00] significance. It is identifiably, visually different. We all know what it looks like.
And when I see that happening, that like little bit of anticipation like, “I did The Thing and now you put food in my bowl” and I love that. I just get such a kick out of it.
Kiki Yablon: It’s like the animal– I mean, that’s just as good an example of what I would call awareness as anything. It’s like the animal testing to see, like what Donahoe said, “is it the ear pricking or is it the metronome?” Like the animal could test that. What if the metronome sounds and I don’t prick my ears? Does the treat still come?
Hannah Branigan: And if the metronome sounded and we paused, I bet that there’s a dog that would go, “… Is it my ears?” And we would say, “Treats!” and he’d say “Yes!”
Kiki Yablon: [00:36:00] A question I have is how much of that is necessary or is any of it necessary for operant conditioning to occur? And I don’t know the answer to that.
Hannah Branigan: How much of like the experimenting? Like, do we have to rule things out on purpose?
Kiki Yablon: Yeah. And I think there’s research on how important is it to have some trial and error and under what conditions and I’m not fluent in that research. I’m sure it’s not black and white.
But on a practical level, I think that we can use these procedures to teach a lot more stuff than we currently use them for.
So there’s certain things that we– And that’s what my presentation was about, ore on a practical level.
So there’s certain things that a lot of us teach that way, like recall and increasingly drop it. Chirag Patel [00:37:00] has like a really popular and great video that I still send to lots of clients where he teaches drop by saying drop and then doing a treat scatter. And then lots of other lovely nuances like I’m also going to put my hand down and point out all the treats so that my hand is not paired with only the moment that you get your shit taken away. My hand is a friend!
That’s something I learned really saliently at Chicken Workshop. Because you know in the Bailey camp, you teach them to discriminate between– You teach the chicken like “peck only the yellow triangle.” And then you start to like fool around with the other things to test them. Like, “if I pick this one up, is the chicken gonna come over and peck it?” And if they do, you would [00:38:00] use negative punishment, which supposedly has like no negative side effects or whatever, you would take away the hot target.
Well guess what? When my chicken pecked the wrong thing, I took away the hot target for whatever, the like 10 seconds they said and then put it back – a little time out from positive reinforcement – and when I started to reach in to mess with the other targets, my chicken started to bite me. I had red marks all over hand from the chicken grabbing and twisting! And I had to fix it by putting my hand out and then offering food. So my hand is not here to punish you. My hand is not here to take the good thing away.
But anyway, I love that part of Chirag’s drop procedure. It’s like having your hand there being a helper [00:39:00] from the very beginning. Like, “Oh look, here’s one over here that you missed.” Instead of like, “Oh, there’s that a-hole hand that’s gonna take away my chicken bone that I just worked so hard to get.”
But there’s things that a lot of us teach that way, and whether we call it classically conditioning the the thing or not, but there are a lot more things that I’ve started to teach that way. It’s easier and I think it’s easier for clients to understand. And so we could talk about some of those.
And then there are some nuances to doing it, I think, that we maybe need to be aware of or like that I’m not sure about yet. But some of the things that I teach that way– And actually I think a lot of people do this–
So you have a great video on teaching your dogs to wait at the door of the baby’s room. And you did that by [00:40:00] basically by stimulus pairing. Like when they approached the door, you clicked, which is sort of an intermediary stimulus that says “look for food” because they might not be looking at you when you if you toss the food without doing that. But then you basically said when you reach the threshold, there will be food behind you. Right? So feed on the threshold as the first stimulus and food behind you is the important second stimulus. And you use the click to kind of clarify it, but I would say it’s still stimulus pairing.
So a lot of people do that for like boundary training. But what I’ve started to do is– So I used to do that and then add a verbal cue and then start to give the verbal cue when there was no threshold for like a wait/hold-up kind of cue. [00:41:00]
Recently what I’ve done is I just start in a not distracting environment, I say “wait,” I pull a treat out of my pouch and present it at my right hip. And I just repeat that and what I start to see is that the dog starts to come to my right hip. It’s basically like a little mini recall. So that’s how I teach wait.
Again, there’s this whole procedure that a lot of us have been taught to teach a dog to go to a mat. Like it’s a shaping exercise that many people’s first shaping experience is “if the dog looks at the mat, click/treat, toss the treat away, dog looks at the mat again, click, toss the treat away. Okay, now you’re looking at it regularly, I’m gonna wait for a step, et cetera,” that procedure.
But I [00:42:00] tend to say the cue or make a cue happen, like opening a door a little bit or touching a doorknob and then I put a treat on the mat regardless of what the dog does. And often the dog surges towards me when I say something or when I get the treat out. But regardless of what they do, this first stimulus predicts that the treat is gonna be on the mat and the behavior of surging towards me for some reason doesn’t usually get reinforced.
That’s one of the nuances is “why not.” And I think you probably could end up reinforcing it if the environment was set up so the dog surged forward every single time. But I just made it so like you surge forward, the treat goes on the mat, you don’t surge forward, the treat goes on the mat, doesn’t matter what you do, there’s no contingency on surging forward. [00:43:00]
Hannah Branigan: Then you may as well go straight to the mat because that’s the most direct path to the reinforcement, which is again, what the system is designed to do for the most part.
Kiki Yablon: Right! I don’t know if this is the right thing to say – I started to look at some literature on this and I just haven’t had time to dig into it a lot. But it seems to me like what’s happening is that there’s two contingencies there.
There’s the stimulus/stimulus contingency and the response/stimulus contingency – the behavior/consequence contingency and then there’s the antecedent/consequence connection.
And if there is no clear contingency on the Behavior, then the A/C kind of takes over!
Hannah Branigan: I’m thinking of when I’ve tried to apply similar techniques when I’ve accidentally [00:44:00] effectively captured not what I wanted and then made it permanent. There was too much of a space between the A and the C, like I was trying to get too much to happen. And where I could see myself getting in trouble, just to borrow from your example, if I wanted to use opening the door– like if I want my dog when I open the front door to go to his mat. If I were to go right to, “I’m gonna just open that door full-on and then put food on the mat,” probably more than half of the time my dog’s gonna be fully in the front yard and before he realizes there’s food on the mat, and so I’m gonna end up shaping, “opening the door, he bolts through the door and then comes back and gets on the mat.” Whereas if I make it smaller and I just put my hand on the doorknob, food on the mat, hand on my doorknob, food on the mat, there’s not gonna be as much space for not-my-target-behavior to happen. And that seems to be more effective? [00:45:00]
Kiki Yablon: I think so, but I have had situations where the animal does go out the door. I think some of it also has to do with “what is the potency of that second stimulus?” Like what is that second stimulus already a cue to do? Me putting food somewhere is a cue to go get that food. That’s already a cue. But if like there’s a bunny outside? Then we have a matching law problem. Right?
But bigger point is that all this is operant. We’re talking about things in operant terms. Which discriminative stimulus is promising more reinforcement.
Hannah Branigan: So even though you don’t have criteria in a way, in terms of if you’re gonna touch the doorknob, “if he doesn’t go to the mat, I’m not gonna feed him.” [00:46:00] I touch the doorknob = I’m gonna drop food.
Now just from a practical side, when you’re doing this, does that dog already have reinforcement history with the mat, with you, like in other contexts?
Kiki Yablon: Sometimes, yes. And sometimes, no.
So one of the situations that I talked about during the presentation was a dog that– When people came up this stairs at this house from– The entry was on a floor below the main floor. And so when people came up the stairs, he was jumping on people at the top of the stairs to say hello, which was dangerous. And I had done like some basic mat training with him, but he did not have the duration and distraction levels necessary for someone to be able to cue him to go to the mat and stay there while the doorbell rang and someone opened the door and the person came all the way up the stairs and opened the gate. That was a lot to ask. I wouldn’t do that. [00:47:00] I would probably just leave the gate there forever.
Hannah Branigan: I have other things to do with my life!
Kiki Yablon: But if you don’t want to leave a gate there forever because you’re bringing your groceries up from the bottom floor or whatever.
But he had a little bit of history with the mat, so when you put a treat on the mat, he was like, “Oh, I know this game!” So what I had them do is just have a treat bucket at the top and bottom of the stairs. Anytime anybody went up or down the stairs, they grabbed a treat and they stuck it through the side of the stair railing onto a mat that was placed there.
I don’t know if the mat made a difference or not, because I didn’t initially try it without the mat. But that basically he started to just see somebody coming up the stairs and he would go to the side of the stairs and then that would be reinforced by the treat. But initially the treat was put out there whether he went there or not.
Hannah Branigan: [00:48:00] That’s interesting!
Kiki Yablon: Probably got a chain, a little sequence where he came to the top of the stairs and looked at the person, and then as the person hit the certain step, moved to the side.
Hannah Branigan: Looked and tail wagging. And then went to the side. Yeah. Which for practical– like that’s functional on a practical level.
So one of the behaviors that I have trained this way for a long time is stand or moving stand, like for obedience, which is just a more formal version of wait in a lot of ways. The end product, I wanna be able to be formally heeling with my dog and then I give a cue and the dog freezes in a standing position while I continue to walk without stopping or pausing.
Kiki Yablon: So you move forward, come back and put a treat where you want the dog to be?
Hannah Branigan: No. So I do it a little [00:49:00] differently. But also kind of the same, so kinda like what you were just saying, which is why I asked that. I have most of the time already built some reinforcement history for standing with that dog before I introduce movement. And it could be just as simple as– Because dogs stand all the time, right? I don’t need to do anything or to do that. So I can just click/treat and click/treat and click/treat. And if I feed like nose level, chest level for the dog, they’re gonna continue to stand and just accept the food. So it’s pretty easy. I don’t need a cue necessarily, but if standing in the presence of a human with food has some reinforcement history, it’s a lot more likely to get it to happen.
And the reason that I often do it that way is because so many dogs, by the time they come to me for obedience, have a history of sitting in that exact same picture because that’s like the first thing we tend to teach young dogs, new dogs. “Oh, I see a human with food! I should sit because that usually works! They love [00:50:00] it when I sit!” And that can make really, really hard to get anything else to happen if they have four years of sitting when a human is in front of them with food. So usually I started it there.
But then what I do– And that’s not on a verbal cue necessarily, Sometimes it is, sometimes it’s not. More and more, because I’m lazy, I don’t bother. I just keep on going. I add the cue at my convenience at a later date.
So then to add movement, I’ll start like walking fairly slowly, small steps, but then I’ll just say “stand” and then peg the dog in the face with a treat. And I say “stand” and I peg the dog in the face with a treat. But if you watch, like if you take the word stand out of it, if you just lift your hand and toss like you’re tossing popcorn for your dog to catch (doesn’t matter if he catches it or not), you’re tossing it, what most dogs do is they’ll shift their weight back and brace their legs [00:51:00] to prepare themselves to goalie that treat. And if I were to toss it like way to the side, you don’t get the same muscle movements happening there. But if you lift that hand like – I’m trying to use different cues for the humans. “Like you’re throwing a dart,” if you lift your hand like that, because of the dog’s learning history with humans in the house, they recognize those kinds of things and most dogs will brace and coil.
And so I can just say “stand” and peg them with the treat and “stand” and peg them with the treat. And then when I say “stand,” they shift their weight back and brace their feet and I peg them with the treat. And so I can capture that in that space. And then I’m walking, I say “stand,” I peg them in the face and I’m walking. It doesn’t matter that I’m walking because that’s no longer the relevant piece of the picture. When I say “stand,” they brace in anticipation of that food. [00:52:00]
Kiki Yablon: There’s a lot of play in there of starting something with stimulus pairing and then beginning to select for behavior.
But I also think what’s interesting is like, “do you have to?” It’s probably more efficient to do that selection thing, but one thing I’m curious about is do we have to?
If we’re teaching errorlessly, you almost don’t need to. Maybe you have that pause in there where you wait and look and see if you see that thing you described with the pony, which I always referred to as thinky face, where they go, “Wait a minute. This?”
Hannah Branigan: I mean, that moment’s so reinforcing for me, I’m unlikely to eliminate it from my training [00:53:00] anytime soon because I just love it. This makes me so happy.
Kiki Yablon: Me too. It’s more of an academic question. I mean that’s usually how I train. So like another example is one thing I talked about in the presentation: thinking about not just the stimulus, but different dimensions of the stimulus, like aspects of it, like in particular location.
So it’s not just a treat. I’m not just pairing the door opening with a treat – I’m pairing the door opening with a treat in a specific place. And same thing with recall. Like recall is not like “I’m pairing the word with a treat.” It’s “I’m pairing the word with a treat at my belt buckle.”
Hannah Branigan: I think that’s so important. Because if you were to open the door and throw a handful of food into the yard, you would absolutely not have a dog going to the mat! That is a critical part of the strategy!
Kiki Yablon: Right. And you might have a happy dog! [00:54:00] You might have a dog that’s wagging its tail. But you’re also gonna have a dog that’s running outside.
So even when doing classical conditioning, we’re thinking about like what are the cue– what are what would be technically called the discriminative properties? Meaning like what is the second stimulus already a cue for? And why?
So like a real common thing I do and I’m sure lots of other people do it too. I just think it’s neat every time it happens so I like to talk about it. If I’m doing like husbandry stuff, like if I’m gonna touch your feet and give you a treat or whatever, or I’m gonna bring the leash to the clip of your harness and then give you a treat, I am going to give the treat where the dog’s head would be if it were turned forward. Which is incompatible with mouthing my hand, biting my [00:55:00] hand, turning in a circle. Lots of one that I don’t want. And so I initially will just be like I’m moving the leash as far towards your harness as I can without you doing any turning. Which if you’re doing classical conditioning, you’re thinking of that as “under threshold.” I think “under threshold” could be interpreted to mean “I’m not getting any operant behavior that I don’t want as well as any respondent behavior that I don’t want.” So I’m not getting increased heart rate and I’m also not getting mouthing, piloerection.
Hannah Branigan: Probably not what you want when you’re putting the harness on, yeah.
Kiki Yablon: So bring the leash towards the harness and then treat, but not just treat, treat upfront. And then when like what you’re talking about, I start to see that when I move the leash towards the harness, I get a head turn towards the front. [00:56:00] Presenting a treat up front is a cue to turn your head toward it so that you can then eat it.
Hannah Branigan: To paraphrase Alex Kurland, treat where the perfect dog would be.
Kiki Yablon: So I think, again, that’s a place where if you think about the operant stuff that’s happening, even if you are thinking of what you’re doing as classical conditioning and the main thing you’re looking for is lowered heart rate, a lot of the things that we look for to tell us if classical conditioning is working are not the behaviors that are influenced by classical conditioning. They’re operant behaviors that we assume mean that some things are happening like inside that we can’t see, like heart rate or hormones or whatever.
Hannah Branigan: Right! Like I loved like several years ago at this point, Lindsay–
Kiki Yablon: Yeah. Lindsay, the head lift.
Hannah Branigan: Her presentation [00:57:00] about, yeah, about conditioned emotional responses. Like what is she looking for when she’s doing her food guarding protocol? It’s the head lift, which is a very operant behavior, like all skeletal muscle. But most importantly, it’s observable without a laboratory.
Kiki Yablon: I mean, I think a lot like respondent behaviors can be observable, so that doesn’t mean it’s operant.
Hannah Branigan: But importantly, I need something that I’m gonna be able to observe in the situation where I’m doing the training. As a dog trainer on the ground trying to train dogs with not enough time, it’s academic to me what their cortisol levels are doing right in the work because I don’t have a way to tell that. So I can’t use that to inform my training plan. Whereas the head lift, that’s a checkbox [00:58:00] behavior for me. Like it is either happening or it is not happening. And I can see it happen on a video. I can see it happen in real time. And I’m always looking for things that I can use in real life, in real time, to tell me “Am I on the right track or do I need to make a change? Do I need to stop and call a friend? What’s happening here?” So that’s what I mean when I say “most importantly it’s observable,” but yes, of course. Like blinking could be both operant or respondent, but–
Kiki Yablon: So a lot of those behaviors that we’re looking at are changeable by the outcomes they produce, which is what we call operant.
And I have an interesting example of this and I think it goes both ways. We’re worried about like, “are we teaching a dog to just give us emotionally empty happy face” or something. I don’t think that’s like a real concern. [00:59:00]
Hannah Branigan: I’ve thought about this because I mean, I socially smile and nod and it doesn’t reflect my private behaviors.
Kiki Yablon: But is a different smile than– Arguably it’s probably very different.
Hannah Branigan: More like a grimace, I don’t know. But I always think that I’m pulling something off and I’m probably not pulling whatever it is off nearly the way that I hope. But thinking like those are probably exceptions.
Because most of the time– Because I got real tangled up on that too and what I kind of thought was, “Well, if I go to the biology of it, all of those responses are there for functional reasons. Like that’s why evolution has kept them.” And so like to grab something like heart rate, which we could measure with the right equipment that’s not out of reach. But yeah, while I [01:00:00] can’t say for certain if my dog is holding still that his heart rate is low, I can say for certain that if he starts moving around a whole lot, his heart rate is going to go higher because the whole purpose of raising the heart rate in that circumstance is to perfuse the tissues, to get oxygen to the muscles. And so if you’re using a lot of muscles, you’re gonna breathe harder and your heart’s gonna beat harder to move more blood more faster to make those movements happen. So it’s a biological requirement for the heart rate to go up.
And if I extend that thinking, well, yeah, that’s actually pretty applicable for a lot of those things that we would put– Why do we salivate when food is coming? Because it’s part of the digestive process that’s necessary for that biological function. So it’s not as abstract as it might seem when I first started pondering on it.
Kiki Yablon: Yeah. I was gonna tell you about– And some people have seen this on [01:01:00] like my social media pages and stuff already. I worked with this dog that was afraid of the ice maker. And the first thing I tried to do was basically just pair like a small, tiny piece of getting ice from the ice maker, like picking up a cup near the ice maker with a treat. I wanted the treat to be delivered in a predictable location so that basically the dog knew where to get it and didn’t have to come towards the ice maker to get it. So the owner was like touching or picking up the cup and then dropping a treat in the dog in a bowl on the floor.
And the dog’s initial fear response was to run down the hall and press herself up against the bathtub in the back of the house. [01:02:00] And what started to happen is owner picks up the cup, dog starts to go down the hall, owner puts treat in the bowl, dog turns around and comes back and gets the treat.
And then what we started to see is the dog standing there and looking at the human and wagging her tail. And so we started to take that as a sort of a start button. So she would start to run down the hall and then she basically would come back when she heard the treat. And then she would start to run down the hall a little bit and come back in anticipation of the treat.
And I was like, you know? And then she’s like standing there wagging her tail. And I was like, I don’t know how to read this body language. Like, we’re partially running [01:03:00] away, so what does that mean? But we’re also coming back quickly and wagging. And what I decided is that I think we’re training her to circle, to do a semi-circle. Because there’s too much space between the stimuli and because the old picking up the cup or whatever is still a cue to do this other behavior. But now maybe she’s doing it as part of the behavior that gets a treat.
And so I had a second person come and be the person who dropped the treat into the bowl. And then we got like sort of a smaller circle. And then I was like, okay.
And then also I was talking to a friend and the friend said, “Is it really a good idea to have tail wagging and looking at the owner be the be the start button? What if you punish it?” I got a little nervous about that, although I think in retrospect I don’t know if I should have worried about that with that particular dog, but I can see that argument.
Hannah Branigan: [01:04:00] Well, I think that’s a reasonable question to ask, but it’s also an easy one to answer. Because if you were to take three data points, is the tail wagging going away or not?
Kiki Yablon: Yeah, good point. So what I did was this dog had no history with a mat, and this was all done over Zoom. So I had the owners do a little teeny bit of relax on a mat, like here’s the mat to get treats on it and the dog was laying down on it.
And then when the dog laid down on the mat, which they didn’t really need for anything else, then that was the start. That was the cue for the owner to move towards the ice maker. And having that little bit of history for “if you stay on the mat, another treat is coming,” it was enough to just compete with that circling behavior. She just moved towards the ice maker and you could see her go like, “Uh, but [01:05:00] I should circle, but also maybe I wanna stay here.” And then the treat came. And caught and not only was paired with the ice maker, but also captured the stillness behavior.
Then we started to deliver the treat so she stood up on the mat so that we could see how fast she was lying down again to get the ice maker to go. Then we gradually increased the ice maker.
So then this is interesting. We took the mat away and the circling came back.
Hannah Branigan: Isn’t behavior wild?
Kiki Yablon: It’s just so fascinating to me. Took the mat away, the circling came back, put the mat back, circling went away. Still getting all the tail wagging. And then they told me, “we use the ice maker the other day while she was in the back of the house and she came running into the kitchen.” [01:06:00]
Hannah Branigan: Awww! “You forgot to put my mat out, but that’s okay. I’ll help you!”
Kiki Yablon: Right? So she just came running in to get her treat. And for me that was the like, “okay, we’ve solved this problem.”
And then what we did to get rid of the circling, which I would say was superstitious at that point, is just faded the mat more slowly. So we faded it into half of its size and then like a washcloth that was the same color. And then gone.
Hannah Branigan: As you’re describing and using the word superstitious, I was thinking about Skinner’s pigeons that spun in the circle– like they didn’t have to. I may be mashing up multiple examples, but he’s delivering food in the hopper and the pigeons just trained themselves to turn in a circle for no reason and that was the classic example of superstitious behavior. And so she’s [01:07:00] turning a circle because that’s just what happened and then she got the treat, and so now “what we do is we go in a circle.”
Kiki Yablon: Yeah. And there’s whole categories of behavior that happen when there’s a delay between things, like when you have to wait for things. There’s debate over whether those behaviors are superstitious, like if they’re happening because they get accidentally reinforced at the end or if they are just induced by that delay. There’s a whole thing you could go into about that. People do all sorts of stuff. They bite their lip, they smoke, they twirl their hair, they pace, you know, and those are sort of behaviors–
Hannah Branigan: “Everything’s terrible. This is taking too long. Oh my God.”
Kiki Yablon: And there’s called adjunctive [01:08:00] behavior. So I think there’s all kinds of considerations when you’re doing stimulus pairing. Even if you’re thinking classical conditioning, “my focus is on teaching a new elicitor for drooling,” there’s other stuff that’s gonna happen in there. And Jesus [Rosalez-Ruiz], this came from Jesus, I think originally. He said, “You know what? Pavlov’s dogs are like strapped into their little harnesses. But what if they weren’t? They probably, in addition to drooling, would also start to run towards the guy in the white coat coming in with the food–”
Hannah Branigan: Wagging and bouncing around!
Kiki Yablon: And it’s possible they were wagging in their little harnesses. I mean, I don’t know. That’s pre-video, right? We just see still pictures. But that’s not the behavior that he was recording. [01:09:00]
Hannah Branigan: He was trying to collect the saliva for his research purposes.
Kiki Yablon: Yeah. He was originally just interested in digestion.
So another thing that Jesus said once that sort of blew my mind – it was like a hallway comment at a Clicker Expo as he does–
Hannah Branigan: Yeah. He does that, yeah.
Kiki Yablon: And I don’t know if he still would say this or not, but probably. He said the main reason to keep dogs under threshold when you’re doing classical conditioning is to avoid reinforcing unwanted operant behavior.
Hannah Branigan: Yeah!
Kiki Yablon: Which is interesting! But then how do we account for like, “Okay, if I just touch the doorknob and throw a treat backwards onto this mat,” how do we account for why that doesn’t reinforce the dog surging forward? And I think we came up with a [01:10:00] hypothesis, which is that there’s not enough of a contingency on it.
But I think you could see times where there is a contingency on it, like with the ice maker dog, where basically every time she circled, she got a treat in her bowl.
Hannah Branigan: Even though it wasn’t deliberate! You weren’t waiting for her to circle, but it was happening consistently enough.
Kiki Yablon: Yeah. How would she know whether there’s a contingency on that or not? It sure seems like it.
Hannah Branigan: So, I do wonder if the function matters. Like, a lot of times when we’re talking about counterconditioning or we’re talking about working under threshold, we’re talking in the context of like aggression or reactivity. So the barking, the lunging, possibly other things, are the problem behaviors we don’t want. But if our training plan– And there’s lots of training methods; this is not the only way to work with these kinds of behaviors! But a common one when we’re talking about this kind of thing is using food, which is totally different function, totally different [01:11:00] stimulus and outcome from probably (in this, my example here), what was maintaining the behavior we’re trying to get rid of, which would be to make the other dog go away.
I wanna work under threshold for a lot of reasons, a lot of reasons. It’s rude to let your dog bark at other people’s dogs. But I wonder if that– I’m thinking like so many things that we’re saying, like, “Well of course, if you threw the food into the yard versus putting it on the mat, of course that would have different outcomes.” That seems like obvious, but also we’re not talking about it necessarily as an operant process in this kind of situation. If you were to say “come” and then pitch the handful of food into the yard– If you were to be counterconditioning your dog and you set it up so every time another dog walks by, you feed him, but you were throwing the food towards the other dog, you would get a different outcome I would expect. It would have a [01:12:00] very different effect than what we normally do, most of us, which is just hand the food or maybe drop it on the ground at the dog’s feet. All of those details matter!
And then I’m wondering like “if I’m not working under threshold, then I’m still putting them in the original contingency, which is making the other dog go away.” So I guess maybe we’re back to Matching Law and then also reinforcing (or not, I don’t know) rewarding, delivering food separately. We probably wouldn’t get the barking to go away because they’re different.
Kiki Yablon: Just some anecdotal observations from my work with Pigeon, my previous dog on reactivity. One is– I mean I tried about a million different things with her. It was like not consistency over the years. [01:13:00] I’m sure you know. I tried counterconditioning, I tried BAT [Behavior Adjustment Training], I tried Look At That [LAT] and ended up doing sort of a blend because I’m not good at following instructions from any one person. And because I think as you learn, when a recipe doesn’t work, you draw on other stuff to make adjustments.
But one thing I noticed was I almost always pulled over to do our work. Right? So you’re walking down the sidewalk and somebody’s coming and my first job is find my working position. And then I’m gonna like hang out there and reinforce for looking and then as the dog is looking, the other dog also goes away without there being any barking which is the ideal there, right? Or I find my working position, we do one repetition of Look At That and then we move further away. Either way, it involved pulling over and [01:14:00] stopping, stepping off of the sidewalk. And so what did I start to see when a person came along the trail? Not just looking at me, but also stopping.
I’ve noticed this a lot in her later years like during Covid, when we were doing a ton of hiking because there was nothing else to do. And I would be like, “There’s a person coming; why are you stopping?” And I was like, “Oh, because that’s what I trained you to do.”
Hannah Branigan: Working with Fig’s car behavior, when I hear a car coming for the bulk of the first part of his training, I would move as far off the road as I could for obvious reasons: getting distance and also not being dragged into the road. And now– Because he’ll hear a car before I can, especially if have ear pods. Now he’ll pull up onto the [01:15:00] shoulder and then look at me. And I love this so much more than spinning and screaming. But it’s pretty fun!
Kiki Yablon: And it’s like a nice environmental safety cue. There’s a car coming, so don’t just look at your owner, like move further away from the car. Beautiful. The car is going to appear–
The other thing I noticed is that, like a lot of people, we did not get in a lot of practice with surprises. If we’re walking down the street and you see a dog coming, we’re fine. But like you are walking down the street around a blind corner and a dog pops out, she would have a reaction. And my way of dealing with that was generally to use a well-trained recall cue and then she would turn around. She would be like, “WOO WOO WOO!” And I’d whistle. And then she would turn around and get treats and we’d move away. And what [01:16:00] happened, not surprisingly, when we came around the corner, is that she would do a short bark and then sometimes stop and look at me.
You can capture that little bit of barking. But the other thing I felt about it was that it was almost perfunctory. So here we’re getting into describing emotions. But I felt like the emotion had gone out of it. And what I mean by that, what that looked like, is it was not as loud, there wasn’t as much piloerection, it could be interrupted easily if I needed to, my adrenaline didn’t go up anymore.
So some parts of that behavior, the operant parts got captured is what I think. But maybe the respondent stuff wasn’t happening as much [01:17:00] because the contingency had changed. Like, now you’re barking to get called away for food instead of barking to make the other dog go away.
Hannah Branigan: Barking in particular really does changes so much when the function is different. We talk about barking as one behavior and it’s super not, anymore than talking is one behavior. So like that’s always been something–
Anyone who’s tried to teach their dog to speak on cue by capturing barking for any other function, as soon as you start feeding, it’s hard to get the bark back. And then when you do get it back, it’s a different behavior. [Delicate hopeful optimistic bark] “Woof? Woof?” [tone of voice: Please? What about now?]
Kiki Yablon: Or what Pigeon would do is take a breath first. [inhale] “Woof?”
Hannah Branigan: Yeah. You get a little like scary hoarse whisper or a cheek puff or all kinds of behaviors happen that aren’t barking. I got a sneeze by accident when I was trying to get [01:18:00] get barking to happen. It sounds completely different than the original more… I would say “emotionally motivated” bark.
Look At That does the same thing. What the behavior looks like when a dog is startling at something that has appeared in the environment looks completely different to that very deliberate head turn with eye roll that they do the third time you click the Look At That.
And I have the same thing with the go sniff. It’s just always fascinating me how the appearance of the behavior changes. And that’s what I take now as how I know I’ve changed the function, which is usually what I’m after in those cases.
Kiki Yablon: Yeah. And then that raises a bunch of questions for me about this sort of newer idea– It’s not really “newer” idea, but, newly unearthed idea of super imposition.
Like, are we changing the function? Are we [01:19:00] superimposing one contingency over another? Is it possible that one happens sometimes and one happens other times and there are factors involved in why that happens? Something I feel like I need to read up more on, what this idea is.
Because the idea is like if your dog is behaving to make a person go away and you teach them an alternative behavior where the reinforcer is something else like food, then you can create (this is my understanding of the description of the problem so jump in if you feel like you have a different understanding), then the problem with superimposition is that when food is not available as a reinforcer, when the new contingency isn’t available, the old contingency is still there and you’re gonna get the old behavior to make the person go away.[01:20:00]
And I feel like I find this to be more true with dogs, but I think there’s a lot of nuance. When your dog’s barking to make another dog go away, I think there’s more nuance to it because of the uncontrolled variables like distance and proximity and what kind of dog and surprise and things like that.
But plenty of times I do think that you can make friends with food, you know? Like you can. It doesn’t end up being superimposition. But I’m fuzzy on what the circumstances are under which superimposition happens and under which it is a problem.
But this is pretty far off the topic of stimulus pairing. Sorry!
Hannah Branigan: Well, I mean, it is, but it isn’t. [01:21:00] So let me see if I’ll drop something in the middle here.
So I’ve talked about this before and I know I’ve talked about it with you. One of the ways that I live with my dogs in my house– I’m out in the country and I am here alone most of the time. So I don’t mind that anyone who comes onto my property knows that I have a lot of big dogs that live here. Totally okay with that. I consider that to be a bonus. That’s part of why I have dogs.
So if someone comes up, whether it’s UPS guy or somebody’s selling something or Jehovah’s Witness, we get those a lot around here too, my dogs are gonna respond the same way and that is to rush the front door barking like the Hounds of the Baskervilles, right? And that’s part of their job.
But because most of the time it is in fact just the [01:22:00] UPS guy dropping off whatever I’ve ordered from Amazon recently, or the Chewy box, could be that. And a lot of times I’m teaching on Zoom or I’m recording a podcasts, so I need some measure of control. And so the way I split the difference is that once they start barking, I throw a handful of food here. It’s not contingent – at least, not the way that I’m– I’m not like waiting for them to come back to deliver it.
Kiki Yablon: It’s contingent on barking!
Hannah Branigan: It’s contingent on barking. But whether I hear the person pulling up first or they do, it doesn’t really matter
Kiki Yablon: Yes. Okay.
Hannah Branigan: I’m gonna throw food here. And because this is happening as part of real life, it’s not a training setup. I’m usually doing something, I’m cooking dinner or I’m answering emails or I’m teaching a lesson. And so it’s very imprecise, but just pitch a handful of food wherever it is that I am in the moment. And so the behavior that emerged was they [01:23:00] all rush the front door barking and then they all immediately rush back here to wherever I am.
And what’s what’s been interesting is I’ve done this for years with now like multiple generations of dogs and it amounts to about the same thing. They all run, they get out 2-3 barks, maybe a few more of they’re fast barking dogs, maybe a few less if they’re Gambit, a big slow-barking dog.
But what’s fascinating is now I could be in the shower. I don’t have food. I don’t have pockets in there. And if someone rings the doorbell or comes up on the porch, they all rush at the front door, they bark two or three times and they run back to the bathroom and wag at me in the shower. And it doesn’t matter that I don’t reward there and I never reward them in the bathroom because I’m indisposed when I’m in that room. But it doesn’t change behavior.
So I found that fascinating. And to tie it back, the barking is different. It [01:24:00] is frequently a pouncier, bouncier– there’s often some tail wagging. It’s a higher pitch barking. And then they run back to the kitchen, the bathroom, wherever it is that I am and then they settle faster.
Kiki Yablon: Yes. Yeah. Recovery time. That was part of what I saw with Pigeon too. Like she could go on with her day after barking instead of like chuffing and straining at the leash of the whole rest of the walk which is what she originally did.
With the barking at the door, it basically tells you when you usually did the first thing that indicated treats were coming, which is after two or three seconds of barking or whatever.
But what’s interesting is sometimes [01:25:00] the barking gets more perfunctory or just fades. So what makes that happen? And I think that is probably, my current hypothesis again is like that it’s whether or not the contingency is relatively consistent, you know? Like if you can get treats in sometimes before there’s any barking a lot of the time then–
Because Pigeon used to– We have like a fenced in backyard privacy fence on all sides except at the end of the gangway facing the street, there’s like a little chain link gate that she could see out of. And if somebody passed by the front of the house, she would like run and hurl herself against that gate snarling.
And this is where I developed– The approach to this is what sort of evolved into the article I wrote about [01:26:00] thanking your dog for barking. I just committed to “I have chicken and I’m gonna follow her to the gate.” I whistle, which is her recall cue, which totally did not work in that situation. And then at that time it was like if I got so much as like a glance I would feed. Now I would probably just feed no matter what. I would whistle and feed.
And what started to happen is she started to not run all the way to the gate eventually. Or she would woof once and start down there and then kind of look at me and I grabbed that. It sort of squeezed the barking out. So sometimes a little barking gets trapped.
Like I’m planning on doing this with Finn because I also am often home by myself and I have had a few experiences where [01:27:00] somebody unwanted came up on the porch and don’t mind having it sound like there’s a big dog here. So I probably will reinforce him for barking once or twice. He currently barks and then like kind of [panting noises], you know, goes back and barks again.
So I’m gonna work on that. But I do want to preserve the barking in his case.
Hannah Branigan: Which is fortunate because completely eliminating it would be super duper – it would be a much longer, more complicated training process than some barking.
And for me, honestly, the most valuable part is how quickly they go back and lay down versus where we would start. Or if I bring a new dog in and they’re not with the program yet, they stay kind of amped up and huffing at the door and maybe running even into the living room so they can look out that window.
But you’re also right because [01:28:00] like I’ve never put any effort into eliminating it because I don’t care. So it’s not a priority for me. And I do care that they do bark. But also just on a practical level, the thing that is most predictable is somebody coming up on the porch or coming into the driveway. And like right now, they’re all asleep behind me because this is our Zoom mode here. I would be able to see someone pull in before them. Because I can see through the window right here and they can’t, because they’re laying down and the windows are higher. Anyways.
But if I’m in the kitchen, I can’t see or hear, so the kitchen’s in the back of the house. And so the way that I’ll know someone has pulled up is because all of the dogs rush out of the kitchen and go barking. And so then I throw food and I walk up. Previously I would walk up and I would like slide against the wall to the point where I could see through the front window to see who was there without anyone being able to see [01:29:00] me. But now of course I have my camera doorbell so I can check that. Oh, it’s much better. I mean, either way it’s not like I was gonna answer the door to be clear.
Kiki Yablon: I think there’s probably also concurrently like a classical conditioning component to all of this too, which is that you’re getting competing preparatory responses. Instead of the respondent’s stuff that’s gonna help you fight, you’re getting the respondent’s stuff that’s gonna help you digest food. And I don’t mean to leave that out too, but I think that’s not usually what gets left out. What usually gets left out is the contingency, the operant and contingency stuff when we talk about this.
Hannah Branigan: It is. That is interesting and maybe this is a good place to kind of [01:30:00] wrap up in conclusion because like I find myself mentally partitioning what I’m doing in a way that’s maybe sometimes a problem for my training. I could do better if I allowed more crossover in my head.
Like as an example, when I would try to– This is why I was kind of laughing to myself when you were saying counterconditioning and I was like, “well, I hope to countercondition. I would like to counter condition.” And sometimes it did and sometimes it did not. And at the time it would seem kind of mysterious as to why it might or might not. And when I started incorporating more Control Unleashed, when I was exposed to Leslie McDevitt and Control Unleashed and her techniques, which she talks about as… I forget if she says “classical conditioning with an operant twist” or “operant condition with a classical twist.” But you know, she’s using the balance of both [01:31:00] at the same time in one training protocol, which is what makes it so powerful.
And I start incorporating the equivalent of Lindsay’s head lift into my counterconditioning. So I’m looking for that little bit of head turn, looking for the eye roll, looking for that little orientation towards me response. And I would say “shaping for that,” but what I mean by shaping is “arranging my A and my C to get that head turn and capitalize on it,” not necessarily making it contingent. So I’m not gonna not-reward because I didn’t get the head turn. I’m rewarding anyways, but I’m going to arrange like where am I presenting the treat, how am I presenting the treat, so that I’m getting a head turn to happen when another dog appears or whatever the thing is.
Kiki Yablon: So if we’re in the summing up phase, I think [01:32:00] there’s sort of two practical takeaways. One is that if you’re looking for operant behavior, there are some easy ways to teach it. It may be easier than what you’ve learned in the past by just thoughtfully pairing stimuli so that the first stimulus proceeds a stimulus that evokes what you want or so that the environment is arranged so that when the first stimulus happens, it’s going to evoke a behavior that the second stimulus will end up capturing.
And then on the flip side, if your focus is counterconditioning or something like that, you still want to be aware of what behaviors are happening, because some of them are your indicators that your classical conditioning is working if you think about it. And then [01:33:00] also because you can get stuff that you don’t want that might lead you to think that it’s not working like the ice maker dog circling. I think a lot of people would look at that and say, “Oh, she’s conflicted.” And yet she runs into the room when the ice maker happens.
Kiki Yablon: And we didn’t even get into how this influences how we deliver reinforcement after a marker signal and all of that stuff.
Hannah Branigan: Yeah, we could do another two hours on that!
I’m seeing it the same way. So my mental partition, which gets me in trouble, is separating if I’m shaping then I’m using operant conditioning, I’m shaping, I’m clicking and treating, then I am sucked into thinking about the criteria for the [01:34:00] behavior and then when describing what we’ve been talking about, this stimulus/stimulus pairing, the question is, “Well, what are you clicking for?” Because it can’t be nothing. But what we’re talking about is in fact that.
Kiki Yablon: We’re talking about like the mat is there and I touch the doorknob and that means that I am going to click in case you’re not looking to tell you that food is moving around and then the food is gonna appear on the mat, and that’s gonna happen when I touch the doorknob, regardless of what you’re doing. Pretty soon you’re gonna see the dog hanging out on that mat usually.
Like the other day I was doing mat training and I was adding distance to a mat behavior, like the dog was laying on the mat and I would like take a few steps away and then I would mark and come back, [01:35:00] or I would take a few steps away and then take a few steps forward, back towards them, mark and come back.
And basically like one time the dog got up before the marker and I still just marked and went back and put a treat where the dog had been laying and on the next rep, which is what’s gonna tell you what effect that had, the dog stayed on the mat, even though theoretically I just reinforced getting up. So a lot of times I feel like it’s more important, especially in the early stages, if the dog gets up, I will still put a treat on the mat because it’s more important to me that the dog understands “when I move away, the treat’s gonna come on the mat” than it is for the dog to understand “you must stay on the mat to get the treat.”
And I think that some data is probably required to see if that’s really as great as I think it is. And I do not have that data. [01:36:00] Except like on a micro level of sort of moment to moment data. If I do that a couple times and the dog starts to get up more, then it’s going south but that’s not– anecdotally that’s not generally what I see. Usually what I see is the dog says “this is wasted effort to get up.”
Hannah Branigan: I kind of go back to how–
Kiki Yablon: Maybe you’re punishing getting up by making them walk all the way back to the mat. Adding effort.
Hannah Branigan: I mean like all things being equal. I think the system defaults to the more efficient strategy. And of course, often in real life things aren’t equal because there’s multiple things going on at the same time, because nature is contiguous.
Kiki Yablon: Right. If they get up because there’s like food on the floor over there, that’s a little different. [01:37:00]
Hannah Branigan: Or their leg is tired or they’re like– There’s so many things that can be going on, that are inevitably going on, at the same time that may or may not be influencing what we see.
But in general, that seems to be in play there.
I think it’s so interesting how intellectually I am very aware and of course completely agree that operant and classical conditioning stuff is happening at the same time all the time. It’s unavoidable. We can’t slice it out. I can talk all day about, “oh, when you’re shaping for operant behavior, of course you’re getting classical behavior at the same time. And when you’re conditioning respondent/classical behavior, you’re getting operant behavior at the same time.” But like really internalizing that and changing what I actually do– To take it from saying that to how I think about it, to how I’m thinking about my training, to what I’m actually doing in the [01:38:00] training, I have to catch myself a lot, even now, like catch my assumptions and say, “Am I really doing what I think that I’m doing? Is there a faster way? And fortunately my system prefers efficiency in this as well. So if I can get the thing to happen by just throwing food, pairing with the thing that’s happening and not having to put a lot more into it, I’m absolutely gonna go that way.
Kiki Yablon: Yeah. And I also think there’s some different precision requirements for some of the stuff you do and some of the stuff I do. I don’t need a specific type of sit or… I’m so far out of sports world that I can’t even think of a good example. I don’t need a heel with the head turned against the front of my thigh or whatever. If things are a little bit sloppy, like the dog takes a step forward and then goes to the mat, like I don’t really care in a pet owner [01:39:00] situation. But I do care about things like that dog circling because that looks like an anxious dog.
Hannah Branigan: No, I think that matters. And the principle’s the same. It’s just my loops are smaller. When I do care about precision, I just have to make that gap that much smaller.
Kiki Yablon: Oh. Yes. And let’s make that connection. The loop is the gap. It’s the the two stimuli and the gap between it and the behavior that gap leaves room for or sets up to happen.
Hannah Branigan: Yeah. That’s an excellent point. That’s a good thing to say out loud. I don’t think we’d explicitly said that. That’s very true.
And so if it’s a big behavior and a certain amount of variability is acceptable for your purposes for it to be functional and solve the problem you’re trying to solve, then you can afford to have a larger loop. And if you– [01:40:00] I mean, you could decide to refine it at a later date. This is a beautiful thing about behaviors. Thank God it’s always dynamic, because otherwise I’d be sunk if I had to get it right the first time!
Kiki Yablon: I’m thinking about some of the situations where my go-to would be “dog goes to a station or a mat,” right? I had a client where we were talking about something else and she was like, “Can you help me with this other thing? I cannot fold any laundry. Like anytime I pick up laundry and start to fold it, like the dog is jumping and grabbing things out of my hands,” right? And we had worked on that with toys, but it hadn’t generalized to laundry.
Hannah Branigan: I can’t fold laundry, but it has nothing to do with–
Kiki Yablon: It hadn’t generalized for either of them. But all I care about there is– And same thing with guests coming in. For the most [01:41:00] part, I don’t care if the dog goes to a station and stays there perfectly until released to greet or whatever. Like I just want four feet on the floor without a mouth on the laundry or the guest or whatever. And so I will basically– That looks like a stimulus pairing procedure because I’ll like pick up a pillowcase and drop food and move the pillowcase this way and drop food. And again, that’s the location. The food is gonna be down low. And I don’t care what behaviors the dog is doing.
Now, you could shape up some weird stuff that way. If you’re not paying attention to what the dog is doing, you might shape up spinning or something that somebody’s gonna call compulsive down the line or excited or whatever. If you make that gap small enough, then you’re gonna mostly catch the dog standing. [01:42:00]
Actually, this kind of goes to something else that we talk about a lot. I want the dog to have many different behaviors they can do when I’m folding laundry. I don’t need them– If I didn’t want a child pulling on the laundry, I wouldn’t put their butt in a chair for the entire thing. I would say “Here is a range of acceptable things that you can do while I am folding laundry for the next 20 minutes.” All of those, uh, will get reinforced. I will talk to you while I fold laundry. We can play some music you like while I fold laundry, whatever.
So eating a Kong, standing, walking around, going and laying down on a dog bed, all of those things are reinforceable. I don’t want precision. I don’t want a little soldier watching me fold laundry. [01:43:00]
Hannah Branigan: So you would just allow– You would create your loop to be only small enough to exclude the behavior you don’t want, but large enough that it could include a lot of other behaviors. Which is again, very much what I am doing, just a lot smaller. It’s just scaled down.
So you know, if I particularly care about how the dog sits, I’m gonna use a setup where it’s very likely to get the piece that I’m after. So maybe I start them on a prop of some sort and then I’m gonna deliver my treat such that if I deliver the treat just ahead of the edge of that prop, the dog leans forward. He’s engaging the specific muscles in his trunk that I’m after for that weight shift. And then he shifts back and I deliver the treat and he goes back and I create– The loop is just occurring over like four inches of space. And yours may be an entire room.
Kiki Yablon: Yeah. [01:44:00] It might be an entire room. Someone else could describe the same procedure that I’m describing by saying that they’re doing classical conditioning to make the dog calm around laundry. So rather than deciding it’s one or the other based on whether we think that the behavior is like an extreme emotion or just annoying or whatever, I think it’s more useful if we just say what we do and get into the specifics of like, “what do you need to do to successfully get the behaviors you want, including the emotional behaviors.”
Hannah Branigan: Whether or not we decide to use a shortcut label, I like the pattern of “I classically conditioned laundry folding, and what I mean by that is I _____.” “Every time I [01:45:00] picked up a sock, I tossed a treat to the dogs mat” or whatever.
Kiki Yablon: I’m about efficiency, so let’s just say what we did.
Hannah Branigan: The system prefers efficiency. Yeah.
Kiki Yablon: Yeah. Especially when you’re talking to clients.
Hannah Branigan: No, that’s a good point. And for me, as soon as I say that I’m classically conditioning XYZ, my head goes to that old thought pattern and I overlook the important operant aspect and I have to watch out for that.
Kiki Yablon: This morning, I was talking about or I was reading in LLA [Living and Learning with Animals] this last session. I think Susan [Friedman] asked me to answer a question about how we talk– Like, does it matter how we talk about what we’re doing to clients or whatever. [01:46:00] And it wasn’t about classical versus operant conditioning or anything. I think it was just a more general discussion of like using technical terms around clients or something or even like using words like cue versus command.
And the point that I made, which she brought back up a few times during the session was that for humans, words are stimuli. They’re cues, they’re reinforcers, they have the ability to influence our behavior. So if you’re thinking about doing classical conditioning, that is going to cue you probably to look at different behavior.
And then this morning I was reading the Washington Post, which had a story by my friend Gretchen Reynolds, who I used to work with actually. She’s a wellness columnist, reports on health science. And she was reporting on some [01:47:00] findings that if people think they are getting enough exercise without changing how much exercise they’re getting, it has health effects.
Hannah Branigan: Oh, that’s creepy.
Kiki Yablon: Isn’t that insane? So the way we talk about stuff is important. Yeah, it does matter and it language is stimuli for us. Very powerful.
Hannah Branigan: This is excellent justification for our hair splitting that we very much enjoy to do.
Awesome. Well, thank you. This was a fun conversation!
Kiki Yablon: Yeah, it was!
[outro]
Hannah Branigan: Thanks for listening. If you like this episode, well, you have good taste and I hope you’ll hit the subscribe button on your podcast app to make sure you don’t miss the next episode. It might be even better than this one.
If you are already subscribed, well thank you. I really appreciate [01:48:00] it, and there are still some ways that you could reinforce me if you were so inclined. You could always leave me a five-star review on iTunes or Stitcher or wherever you happen to be listening to this podcast. And you can also check out and support the sponsors because they help make the podcast possible. You can find links and information about them and the other things that we’ve talked about in this episode by going to the show notes, which can be found at www.wonderpupstraining.com/podcast. And while you’re there, you could also pick up a free PDF training template to help you plan your training sessions. There’s also some other articles and previous podcasts and that sort of thing, which you could always find if you were interested.
So until next time, happy training!
