How To Train a Dog To Recall Through Building Foundations
Do you dream of the day that your dog reliably comes (“recalls”) when you call them? Or maybe you’d just settle for an improvement relative to where you are right now. I recently decided that the best thing I could do to improve my little hunty-sniffy dog’s recall was to start over, and I’m bringing you along for the ride!
Navigate to Where You Want To Go
In my previous article, I talked about what recall actually is, why I changed Sully’s recall cue, and how I started working with the new cue (among other things). While I plan to systematically work on her response to that new cue (“Ewok”), today, I want to talk about the training that I’m doing simultaneously that doesn’t involve any verbal cues at all. We’re going back to foundations! (Pssst … I cannot emphasize enough how much the training in our on-demand video e-course, Attention Unlocked, can help with recall goals).
Let me start by painting a picture for you of my two dogs’, Otis and Sully, typically offered behaviors on trail walks (aka what they do voluntarily rather than in response to a verbal cue). As you read, consider which dog’s history may make them more likely to recall and why.
This is a pretty accurate depiction of the different behavioral tendencies of my two dogs. There is no good vs. bad or right vs. wrong here - just different. These dogs are individuals with vastly different learning histories. I love that Sully loves nature so much and my goal is not to turn her into Otis.
OTIS: Otis trots along trails sniffing specific spots or air scenting (aka throwing his nose in the air and following some invisible scent through space) and tends to stay in a roughly 15 yard radius from me. If he hits the edge of that radius, he either voluntarily stops and looks back at me or runs all the way to me, which I usually reinforce with a treat. As he trots along within that radius, he regularly looks back over his shoulder to see where I am or returns all the way to me to check-in, which I tend to reinforce with a small treat. If he sees another dog or person on the trail, he automatically stops and orients to me, which I tend to mark and reinforce. If I stop walking, he stops walking and looks at me. If I continue to stand still, he runs to me. If I change directions, so does he. If he hears a treat bag rustle or sees my hand move towards my treat bag, he sprints to me. If I give Sully a treat, Otis shows up.
SULLY: Sully moves with her nose glued to the ground for the vast majority of her walk. She has gone entire trail walks without ever once looking up at me (no joke). If I stop ahead of her in the middle of the trail, she will arc around my legs and carry on sniffing as if I am a rock and she is the river that just flows around it. She has been known to walk right past treats offered to her in open palms early on in walks. When I give Otis a treat, she appears oblivious by the way she carries on walking and sniffing right past the whole event. If she is not sniffing, she is likely staring at some prey she saw or heard.
Chances are you determined that Otis is likely better positioned (at least based on the limited info I gave you) to respond to recall cues on trails than Sully. How did you conclude that since I didn't talk at all about calling them to come? You probably picked up on the fact that Otis is already regularly offering components of the recall behavior on walks, which means those behaviors are getting regularly reinforced in that context. Plus, if my dog is behaving to access my reinforcers, I may have a bit more confidence in my ability to actually reinforce recall in the future (clearly, I needed to make adjustments with Sully).
This is a clip of my two dogs, Otis (gray and white) and Sully (blonde), and my neighbor's dog, Chance (white), on a hike. There were clearly some interesting smells in that area that they all picked up on, which thrills me! I want them to have the chance to move and explore without focusing on me. But notice how Otis voluntarily runs to me after he’s done sniffing? The environment cues those check-ins all the time with him. With Otis, I also feel confident saying that my treats compete easily with most of nature most of the time since he often behaves to access them. This type of history is really helpful when it comes to recall!
“Attention” (i.e. orienting to you) is one of the most important foundations for recall. I want to see a dog offering attention in the environments where I will use their recall cue in the future. Ultimately, it’s easier to recall a dog who is attuned to you to some degree (even if it’s just an occasional glance back) than one who has all of their senses fully tuned to the environment all the time. But perhaps more importantly, attention is the first behavior in the chain of behaviors that make up recall.
When you call your dog, they have to stop (if they are moving) and orient to you before they ever run to you (that is basically how we define “attention”).
I put a lot of weight (in the form of a positive reinforcement history) behind offered attention because I want that behavior to show up more often outdoors. By focusing on offered attention (aka no verbal prompt), I can reinforce part(s) of their recall behavior while mitigating the risk of using their recall cue when they’re not capable of responding, which would just weaken it. Plus, depending on when I mark and how/where I reinforce, I can actually produce more of the full recall behavior. For example, if my dog looks at me when they are 10 yards away from me on the trail, and I say “yes,” that marker cue essentially pulls my dog all the way into me. From the outside, it will look just like recall!
There is another foundation that is often overlooked. It’s easy to focus so much on the recall behavior that you take for granted the behaviors needed to actually access your reinforcers. Imagine you call your dog, and they run to you. When they get to you, you offer a treat in your palm. Your dog looks at it and decides to return to sniffing instead of eating it. You were hoping to reinforce their recall with a treat (we might say that “giving them a treat to take from your hand” was your “reinforcement strategy”), but they have to perform behaviors (approach hand, open mouth, grab treat, swallow treat) to actually “get the treat.” This leads to the other foundation I want to highlight in this article: Your dog needs to reliably (in a variety of environments) respond to the cues and perform the behaviors associated with the reinforcement strategies you plan to use with your dog’s recall. Put actionably, you likely want to practice your reinforcement strategies (in a variety of environments) before trying to use them to reinforce behaviors. Now put more plainly … think about how you plan to reinforce your dog’s recall – for example, dropped treats, tossed toy, etc. – and practice having your dog eat dropped treats or chase a dropped toy in various environments without those valuable events being contingent on some behavior like recall. If your dog doesn’t reliably eat food on trails, then you won’t really be able to use treats reliably to reinforce recall.
Markers are conditioned reinforcers, but they are also cues. That means if we plan to use markers, which I recommend as a general practice in training, we need to make sure our dogs can perform the behavior that the marker cues to actually access the reinforcer. In this video, I marked Sully’s offered attention by saying “yes,” which should tell her to come and get a treat from me. The first time I said “yes,” she started to move but then stopped. The second time I said “yes,” she flew to me to get the treat that “yes” told her was available. Ideally, I don’t ever want to use my marker without it being followed up by the associated primary reinforcer (this pairing needs to remain tight and consistent), so I want to be careful about only using it in conditions where I am confident that my dog can actually respond to get the treat (or whatever reinforcer my marker signals). I am way more careful about when I use my marker with Sully now.
Your reinforcement strategies may or may not include the use of markers like the word “yes” or a click from a clicker. Markers are a fairly complex topic that I am not going to dive into too deeply here, but feel free to check out this Instagram post I did talking about markers. I bring this up because strong markers allow you to capture offered attention from a greater distance. The marker serves as a secondary reinforcer for the offered attention (i.e. looking back at you) and a cue that brings your dog all the way back to you to access a primary reinforcer (i.e. a treat). However, in order to do this, not only does your dog need to be able to perform the eating behavior, they need to reliably respond to your markers. With some dogs, they’ll almost “automatically” respond to their known markers in just about every environment, but for many dogs, you have to train your markers in new environments to get your dog to respond to them there.
It’s easy to think that recall training is all about calling our dogs, them coming, and then reinforcing that behavior. But there are a lot of other things we can do that can improve their recall! For example, just about any training you do is going to up the level of reinforcement your dog has for working with you, being near you, responding to you, etc., and that history can absolutely help make it more likely for them to come to you when you call them.
I am using a few strategies right now to get Sully to offer me more attention on our trail walks, but I could boil them down to this: I’m reinforcing the heck out of it! Here’s a closer look at what I’ve been doing (you can see some of these in action in this IG post we shared):
1. We pause once on our daily walks to play the up-down pattern game (popularized by Leslie McDevvit) for 30 seconds or so. I love this game because it’s simple, sets the stage for her to offer attention (the pattern helps), and allows me to reinforce a lot of reps of offered attention pretty quickly!
To play:
Place a treat down on the ground in front of your dog’s paws for them to eat.
When they eat it, they should naturally lift their head up. Mark (you can say “yes” or use a location specific marker like “find it”) and reinforce (place another treat down in the same spot in front of your dog’s paws.
Quietly wait for them to orient up towards you. (Many dogs will look up at you, but if that is uncomfortable for your dog, general orientation towards you works just as well!)
When they do look/orient to you, mark and reinforce with another treat in the same spot. Keep repeating.
Tips:
This game is about capturing offered attention, so you want to avoid prompting your dog by saying their name or pointing to your eyes.
Your dog is allowed to look around. Just wait for them to look up.
If they aren’t looking up at you consistently, you may need to start in a lower distraction environment.
2. We pause on walks for 30 - 60 seconds to play moving pattern games. These games give Sully a chance to practice actually running to me.
To play:
Place a couple treats down on the ground and then move away (if you’re just starting, take only a few steps, but if your dog knows this, you can move farther away).
After your dog eats the treats you put on the ground, they will look at you. Mark and reinforce by putting treats down on the ground where you’re standing and then move away again.
Repeat.
Keep it small or make it a bigger game with more running!
Tips:
Just like in the up-down pattern game, you are capturing offered behavior rather than verbally prompting your dog to come to you (though you could adapt the game to that end in the future!) In other words, do not say your dog’s name or their recall cue (for now).
3. We capture any and all offered attention as we walk. That means literally anytime Sully intentionally chooses to orient towards us, we mark and give a treat (or just give a treat). This isn’t a behavior that has been showing up much on walks, but we are hoping to see more of it after adding the three items on this list to our walking routine.
We actually made a whole on-demand video e-course called Attention Unlocked with Juliana DeWillems of JW Dog Training to teach you how to build attention from the ground up (you would start farther back than what I am talking about in this article). Attention may not sound that sexy, but it’s the foundation for so many other behaviors (including recall). You will often reap big rewards by spending time on foundations (even if it feels “easy”) - we truly cannot recommend it enough.
The three things I am doing right now are very low lift, which is key if I want to reliably do them. More broadly speaking, these activities have dramatically increased my reinforcement level with Sully. One of the indirect factors that often helps a dog’s recall is just a large R+ history for interacting with that person, being near that person, etc. I want being near me to predict GREAT things for Sully, and since dogs are always learning, even basically any training can improve recall (this is why so many people who did the TOC Challenge reported that their dog was coming when called so much better even though there is no recall in that course).
Right now, I am just trying to get SOME offered attention on walks. In the future, I may incorporate some stimulus-stimulus pairing procedures to help produce more offered attention at certain points on our walks. For example, I may drop treats anytime I stop walking (pairing my stopping with treats on the ground) to try to produce the operant behavior of her standing near me. Or maybe I’ll start dropping treats anytime she hears a squirrel moving to get squirrels to become a cue to orient to me. Or maybe I’ll pull a page from Attention Unlocked book and work on offered attention around specific distractions. We shall see … for now, we are starting nice and simple and just trying to get attention to show up in the context of the woods!
Do you have a dog who will wake up from a deep sleep and come over to you if you even open the treat drawer? This is very much related to recall! One day I realized that anytime I opened a certain type of Tupperware, both of my dogs showed up in front of me. This is a great example of operant behavior (walking to me) that was produced by stimulus-stimulus pairing. Now to be honest, I am not sure where the pairing actually started. Was it the smell of chicken that got paired with me handing them chicken? Was it the sound of the lid that got paired with the smell of food, which might have already had its own associated response? Was it the sound of the lid that got paired with me handing them food? At any rate, the sound of that lid became a cue that signaled to them that if they came to me, I would hand them food. This is a reminder that our dogs are always learning by consequence and by association. The way to build a recall cue that’s as strong as the sound of treats or Tupperware lids is to make our recall cues reliable predictors of GREAT things.
We know that dogs perform behaviors more often if they lead to valued reinforcers. But the challenge is that the value of a stimulus or event is not static.
Let’s look at a human example. Imagine you got lost hiking in the woods and walked eight miles without any food. How valuable might a giant burger be to you in that moment? Now imagine you just ate a huge brunch and had to unbutton your pants to give your belly some room. How valuable might that exact same burger be in this post-brunch moment? Is it as valuable as it was to you after your full day hike?
I’m talking about motivating operations (MOs). “Motivating operations influence the current value of a consequence and therefore the frequency of the behavior that would contact that consequence” (The Dog Behavior Institute). DBI has a great post on MOs and one on the difference between a motivating operation and a discriminative stimulus if you want to learn more.
A few people have asked me how I chose “Ewok” as Sully’s new recall cue. In terms of criteria, I wanted the cue to: 1) Delight me (listed as number 1 for a reason); 2) Be something distinct that didn’t get used regularly in everyday life (really didn’t want to dilute it); 3) Be easy to say (aka no words that I might trip over trying to say); 4) Avoid causing harm to people when yelled out loud (i.e. don’t want to be yelling something that makes other people uncomfortable). Plus, before we got Sully’s DNA results, we assumed she was part Ewok. We still kind of think she is.
With Sully, the value of my treats in a given moment influences how likely she is to orient to me (or recall in the future) since that’s the behavior that leads to treats. I want to do everything I can to up the value of those treats (well, not everything … I am not going to use deprivation and starve Sully before training).
I play with MOs a bit to help myself out (and to be kind and care for Sully’s needs). We’ve been doing most of our attention training on our morning trail walks, and I let her walk for at least 20 minutes without any interruption from me. If she happens to offer me attention, I may** reinforce with a treat, but I am not going to pause to play games until after that 20 minute mark. I want her to “get her fill” of nature first so my treats might go up in value a bit relative to nature.
Imagine you’ve been sick and trapped indoors on the same couch for five days, and you feel like you’re going to lose it if you don’t breathe some fresh air. Getting outside has become REALLY valuable. Now think about Sully who has been “trapped” indoors all night as we sleep and desperately wants some nature in the morning. She’s been “deprived” of nature for the night and it’s value is quite high first thing in the morning, which means my treats are likely lower value relative to it. The value of my treats go up as she spends some time in nature and isn’t feeling so deprived of time outdoors.
**The reason I said “may” is because Sully’s eating behavior is a bit more fragile than many other dogs I work with. I had to work hard to get her eating behavior to show up consistently outdoors, and a big part of our success stems from my not offering her food when she’s likely to refuse it (I don’t want her to rehearse that behavior). In the first part of a walk, the environment is so high value that it’s hit or miss whether she’ll eat. I make a decision whether to drop a treat or not based on what exact behavior I’m seeing in her offered attention and what’s going on in the environment. After we get about 15 minutes in, it’s usually pretty safe to assume she’ll eat.
Data!!!!!
Sully’s baseline for offered attention or check-ins on walks was basically zero. If I am actually reinforcing attention, I should see more of it under similar conditions moving forward.
My dog Otis regularly comes up to me on walks “asking for treats.” I want to see some of that from Sully to feel more confident that I’ve got a reinforcer that can actually do some reinforcing!
Here’s the good news: She offered us (me or my partner, Ben) attention on this morning’s walks FIVE times. That’s a 5x increase given that we started at zero. (Note: This count does not include instances of attention that are a part of the structured games we play.) AND, she even turned away from something in the environment and voluntarily ran to me (this was at the end of the video I shared on Instagram). This could be a fluke, but I don’t hate what I’m seeing … !
I want to be clear that I am not trying to turn Sully into a dog who is always focused on me. I love Sully’s love of nature. I am constantly checking myself as I train to make sure I am not depriving her of what she needs to thrive (time to “independently” explore nature being a key part of that). She never has to “earn” the ability to sniff or move around, and I still want the bulk of the walk to be hers. I am just seeing if I can get a few more tiny moments where she connects with me and then returns to the rest of nature.
More to come!