What if your dog’s recall is so good that she comes before you call her?
The little movie featured in this post shows the myriad ways a smart dog can mess up your plans.
The Original Goal: Film That Recall!
This post brings together a lot of ideas I like to explore. Among them are cues, offered behaviors, and stimulus control. And like many of my posts, this one features an unintended consequence. I’ll explain.
Zani has a great recall. Bragging material. Even though she is getting older and her gait is wonky, she’s a nice example of a hound who comes when she is called.
I’ve been wanting for a while to get a video or two of this “miracle” of training. The most impressive recall videos demonstrate low latency (a speedy response) while dealing with high distraction. No problem, I think. Zani’s recall is that strong. I’ve been calling her out of play, away from other dogs, and away from varmints for years. I can film that.
So when Zani fails to come indoors when Clara and I do, it seems like the perfect opportunity. She’s involved with something in the yard. I think, “Aha! I can call her with her “real” recall cue and record it!”
But I don’t have any high-value treats, so I’ll need to go into the house and come back with some. Don’t misunderstand; the food isn’t a bribe. I’m not going to stand out there and wave food at her to get her to come. She will respond to her verbal recall cue whatever the circumstances. I’m getting the food to keep the reinforcement topped up so she’ll also come the next time.
Here’s the plan.
- Zani is involved in something in the yard and doesn’t come in the house when first invited. (I often give her a choice about coming in or staying out.)
- Eileen goes into the house to get a great treat.
- Eileen comes back outside. She has her phone ready to take video.
- Eileen calls Zani.
- Zani comes with Eileen filming.
- Zani gets the great treat.
And that’s how it worked for the first couple times. But the videos had flaws that made them unusable. I filmed poorly, or she slipped. Something went wrong. But I kept at it.
The Problem: Dogs Notice Patterns
Dogs are known for paying attention to us. In my experience, dogs who can earn treats for behavior around the house get very attentive.
Zani is a great example of this. She is a brilliant little problem solver. Let’s rewrite my list of events from Zani’s point of view.
- I’m barking in the yard.
- Eileen goes into the house.
- Eileen comes back out holding the camera.
- Eileen calls me.
- I go to her.
- I get something great.
Question: What is the immediate predictor that a recall will be reinforced with something great?
Answer: Eileen gives the verbal recall cue.
Question: But what happens before that?
Answer: Eileen comes out the back door holding up her phone.
Bingo. Why should Zani wait for me to call her when she already has an excellent predictor of reinforcement?
It didn’t take but a couple times for Zani to learn the pattern. She started running toward me as soon as I came out into the yard holding my phone. Before I called her. She ruined my movie! I’m supposed to call her before she comes. Otherwise, it can look like she is randomly running to me.
Now, it’s moderately interesting that one of Zani’s recall cues is me coming out the door holding a phone up. But there’s nothing earthshaking about it. It’s behavior science 101. All creatures with a nervous system (and maybe some without) are wired to notice predictors. Antecedents and consequences drive behavior.
The really interesting thing about this is the human response. We are biased toward cues that we give deliberately, especially verbal ones. Some of us refer to them as commands, as if they were inviolable. We think of environmental cues as somehow less important or less real. But here’s a hint: dogs probably don’t.
As much as I might get frustrated with Zani’s strangely cued recall, it’s just as impressive as if I called her. My little hound will turn away from something that is very exciting in order to run to me for the likelihood of reinforcement. But it doesn’t feel as potent or valid because I’m not verbally cueing her or using hand signal. That’s so silly, but that’s a human for you.
I’ve had several epiphanies about behavior science over the years. The first and biggest one was about reinforcement. But the next one was about cues. Do you remember the first time you heard someone use a cue that wasn’t a description (in English or another language) for the behavior? Did it startle you a little? It did me. “Wait, you can’t use that word because the dog doesn’t underst- ….oh.” That was the first step for me to get that dogs don’t understand language the way we do. Despite interesting research about areas in their brains that correlate with ours, we can’t assume that they understand word meanings or syntax. Humans always hear the meaning of the word behind the sounds. In fact, we can’t un-hear the meanings. But it is still only fair to assume that dogs have to brute-memorize the sounds.
So to them, there may not be much of a difference between “see Eileen come out the door with a phone” and “hear Eileen make a certain sound.”
Should there be a difference to me?
I’ve written recently about my lack of focus on stimulus control. I’m a bit self-deprecating about it, but it is a conscious choice. In most cases, the behaviors I ask my dogs to perform are useful even if I haven’t asked for them. But I do think it through in each case.
You’ll notice in the movie that I reinforce Zani for coming to me even though I haven’t called her. That was a decision point. I could have discouraged that developing recall cue by not reinforcing Zani for coming unless I had actually called. Instead, I considered: Is it a good thing for my dog to interrupt what she’s doing because something in the environment suggests that it’s a great time to run to me? Is there a reason that might not be a good thing?
In my situation, it’s almost always a very good thing. So I reinforced it. But there is a circumstance in which stimulus control on a recall would be desirable. That would be in some kind of accidental situation where my dog was loose and ended up on the other side of a busy street or another hazard from me. If my dog comes running to me when she sees me appear and there is something dangerous between us, she could get hurt.
As much as we like 100% guarantees, we rarely get them in life. I weighed the probability of that dangerous situation against the benefits of a dog who comes “before I call her.” In my case, the default recall provides enough benefits that I’ll take that small risk. I can also train a different behavior that I can cue during a rare emergency. But for me, it’s great to have a dog who will turn away from exciting things to check in or come to me on her own. Others who take their dogs into lots of situations off leash might decide differently.
By the way, I trained Zani’s recall by 1) giving her something fabulous every time she recalled on cue; and 2) giving her something good-to-fabulous when she came to me in many other situations. Pretty straightforward. That’s why I love training recalls.
How about you? Do you reinforce “offered” recalls? Got any interesting cues, deliberate or accidental?
- Doesn’t Intermittent Reinforcement Create a Stronger Behavior?
- Calling My Dog off Rabbit Scent at Night
- Safety Behaviors: Down at a Distance and Recalls
- Stimulus Control, Or Why Are There Seven Shoes on the Table?
- Stimulus Control, Or Lack Thereof
Thanks to Erin Topp for helpful suggestions about this post.
Copyright 2019 Eileen Anderson
26 thoughts on “Oops! I Trained the Better Than Perfect Recall!”
Whenever I call my 10 year old dog in from the yard she gets a treat when she comes in. She also gets a treat when we return from a walk. In the last 6 months or so, she has made a creative leap. So far I have been unable to understand what is behind the change. Whereas before she knew a treat was coming when she came in the house, she has now decided that the ‘coming in the house’ behaviour does not have to be done by her. Now whenever anyone comes in the house she goes and sits expectantly by the cookie jar. It’s hilarious but puzzling.
Oh wow, that’s an interesting one! Have you fallen for it and given her one? (I totally would.)
Nope! Sitting by the cookie jar is cute, but so far I’ve resisted treating this new behaviour.
One unintentional cue for my dog is me reaching into my bag. Where we like to walk is a wooded trail, and there’s a few places where it opens up and there’s great big fields on either side. I usually stop and throw the frisbee for her a few times at one or the other of these open spots. Which means I stop, and then reach into my bag for the frisbee. She’s usually off leash up ahead of me, but as soon as she sees me stop and reach into my bag, she comes running back to me, ready to play. The thing is, sometimes I stop and reach into my bag for my phone. That gets the exact same reaction. Bag = play time for her.
That’s neat. I love how observant they are. (I love it most of the time, anyway!)
Just be sure not to throw your phone….
I have no doubt that she would chase after it and grab it in her mouth. She’s not shy about taking objects of different sizes and textures. XD
Working with Ray in general produced some interesting training notes:
1. Never teach “OK” as a release instruction. It is simply too common a term and total strangers will inadvertently release your dog from its last position!
2. They read body language much better than us, and it will over-ride a verbal instruction as necessary. To anybody who thinks they have the dog trained with verbal instructions, I suggest they turn their back to him/her and say “sit”. The end result will determine whether you did as good a job as you think!
3. Do not delay the reward as dogs live in “the moment”. i.e. Telling Ray to sit when he is 30ft away is fine, but then don’t call him over to reward him for it, as he will believe the reward was for coming over to you.
4. Don’t keep repeating the instruction until you get a response. Saying “Sit” five times before he sits could well be teaching him to sit after hearing it five times! They are very literal.
5. As for recalls? Ray’s recall is inversely proportional to the small animal activity in the area. i.e. He is very good, unless he sees a cat … and then we don’t exist!
Working with dogs is great fun and can be so rewarding, but I get annoyed at those individuals who seem to believe that dog training is intuitive. It isn’t!
I have an accidental cue. For two years I have used , Piper come or just come. I’d say it works 90-95% of the time. Then one day a few months back Piper was at the back of our yard and there was a deer at the fence. Piper wasn’t barking just looking , I said her name and she turned to look at me and then I said come…but she just stood there…so just by accident I called out Search and she came running at full speed down the yard. I had been using that word for when she does a behavior and I throw a treat for her and I have used that word thousands of times over the the last two and a half years and she has a gotten a reward every time she hears it. I use it now for treats thrown on the ground or to come to my hand and get a treat. People who hear me use it (like the judge at her first PWD water trial in Sept) seem puzzled till I tell them the story. Unusual, maybe, but I will continue to use it!
It’s so true what you are saying about our human valuation of different words or signals. Many years ago, I had a young woman in one of my beginner dog training classes with the typical beagle that just wanted to sniff the ground and not pay attention to her. After a couple of weeks, she comes back to class sad, thinking her dog is not doing well. She explains that her beagle will come running to her no matter what he’s doing but she has to call out the word, “Food” instead of, “Come.” She was happy again when I reminded her that dogs don’t speak English and whatever she did right with the word “Food” could be done with “Come” or she could just enjoy her great “Food” training!
For a long time, I assumed that a fellow dog-walker’s dog was called ‘biscuit’!
I go back and forth on this! I like to reinforce checkins on my hikes, but if I’m not careful, I end up with a dog who won’t leave my side. Barley is pretty fetch-obsessed, so if I reward him for coming to me with a stick throw, he won’t do ANYTHING else on our hike. Not a huge deal, but I like to encourage him to go smell the roses a bit. We have to use an “all done” cue to get him to leave me alone some days.
I have a neat one. Several years back, I unintentionally trained my dogs to be quiet when I told them ‘shh’.
I used to live in a rural area (I miss being in the woods) and I always stayed alert when out and about with the dogs just in case a wild animal was nearby.
If I heard a noise such as coyotes howling or the telltale snapping of sticks that indicated the presence of a black bear, I would freeze to listen and tell the dogs ‘ssh’.
The dogs soon realized I only made that sound when there was something to listen for, so now unto this day when I ‘shush’ them, they stop all movement and noise, even panting, to pause and listen with me.
It was completely accidental, but I soon realized it was a fairly valuable cue.
“You stinker!” LOL Brilliant observations. Don’t you find it’s amazing how quickly stimulus control can be transferred sometimes?
Tilney has learned that setting up the camera tripod = training, and he’ll run over to his usual training area and start bouncing around in excitement. It’s very cute. 🙂
I do reward him for offered recalls. He’s a terrier, so I’m just trying to build that reinforcement history as high as I possibly can. I think of it as “check ins”—almost any time Tilney chooses to leave the environment and come say hi to me, he gets a reward.
Thinking about it, I also wouldn’t be too worried about the situation you describe because one of his strongest cues is “wait!” I originally taught it for doorways, but pretty early on I was hiking with him and he scrambled after my dad down a steep drop and almost pulled me down after him. I instinctively called “wait!” while I regained my balance, and to my surprise, he stopped perfectly! Now I use it all the time, in all sorts of contexts, so I feel that in an emergency, it’s probably what I would be most likely to yell, and something he would to listen to.
Difficult the choose from the thousands of examples…. My favorite was Elly, my Lab, the first dog I trained for Utility. She did nice go-outs, but found it *very* difficult to wait for the verbal cue/hand signal for the directed jumping. If I moved my eyes (not my head) towards a jump she would self-release, jump, and come soaring in for a front. Mind you, she was sitting 40 feet away, and how she could she such a tiny movement beat the **** out of me. I had my friends stare at me to see if I was giving off any other signals, but the no one could see anything but the eye movement.
I’ve got a good one. I unintentionally trained my dogs to be quiet when I told them ‘shh’.
I used to live in a rural area (I miss being in the woods) and I always stayed alert when out and about with the dogs just in case a wild animal was nearby.
If I heard a noise such as coyotes howling or the telltale snapping of sticks that indicated the presence of a black bear, I would freeze and tell them ‘ssh’ while listening.
The dogs very soon realized I only made that sound when there was something to listen for, so now unto this day when I ‘shush’ them, they stop all movement and noise, even panting to pause and listen with me.
It was completely accidental, but I soon realized it was actually a very useful cue.
Sorry you had to submit your comment twice, but I like both versions, so here they are!
Great post, and great to see Zani so full of pep!
My Piper offers “recalls” quite a bit. And we made the choice to reinforce them, too. She is quite independent, so any opportunity to reward her presence to us is welcome, in my book. We also want her to know that coming back to us, for any reason at all, is always the right answer. She tends to be a very social dog, so getting recall when she was younger when other dogs or people showed up on hiking trails was a challenge (she almost mugged Kasich, yes the Governor of Ohio, when we were hiking off-leash last year, but we managed to call her back before she saw him and his security detail on the trail). What we’ve found is that the whistle is almost 100% reliable, much more so than verbal. I think this has a lot to do with the humans blabbing so much when we hike and the fact that the whistle is so unemotional when blown. It always sounds the same. And it’s always rewarded. We have called our dog in hot pursuit of deer at 200 yards – where she comes running back to us just as hard as she was after the deer, just so she can get her cheese.
Oh, yes. A commercial break starting or a movie ending (I suppose the key here is actually my disengagement, change in posture, etc.), the sound of a laptop lid closing, the sound of the kitchen scale being lifted off the counter, the sound of the treat jar being moved, the sound of a plastic bag or container with treats being lifted from the fridge, the sound of food being chopped, the sound of the clicker being lifted from the shelf, my approach to the corner of dog things (toys and grooming stuff) and I’m sure some other signals have at some point served as reliable recall cues in our house. 🙂 I don’t reward my dog for this specifically, but of course a walk or dinner or training session has followed enough times to make it stick. Nothing gets past her! I often reward her for offered check-ins and recalls on leash walks and always always always when off-leash.
I would have reinforced as well even if the dog responded before I gave my verbal cue. Interestingly, I had the hypothetical “what if my dog is loose…. situation once with our Aussie. She did the very rare thing of jumping out the car before released, ran across the street to the park, realized her mistake – I suppose – and was about to return to me when a car approached. She did have a very solid wait on cue as well, and with that I could stop her. She stayed where she was at until I was able to cross the street, and reinforcement the heck out of her for staying, even though she had jumped out of the car
On the topic of how humans value some cues more than others, in dog training classes I have been puzzled as to why I am supposed to phase out my non-verbal cues and replace them with supposedly superior verbal cues. I can see the point of some cues needing to be verbal but dogs do seem to respond well to non-verbal cues so why not capitalise on that?
Verbal cues are invaluable if your dog is not looking at you! Non-verbal are really important but do have obvious limitations. An ideal scenario is a good resource of both verbal and non-verbal. 🙂
People think our Ray is called Buddy. He responds to both! 🙂
I’ve been experimenting with an idea. I use a different marker word “Goodie” when I know I am going to give my dog a high value treat (not just for recall but for other behaviors she finds difficult). You see I loved the idea of the “Bacon” recall I have heard about. But I found I wasn’t bothering to invest the time to train it on it’s own, so I figured maybe if my dog gets used to turning to me for her high value treat after other behaviors such as staying on a loose leash with a squirrel in sight, maybe that marker word can then become my emergency recall maintained by an association that is simple to maintain in everyday life. She does have the typical recall word “Come” delibrately trained.
Non-verbal cues are invaluable if your dog goes deaf. And as ColinandRay points out, verbal works when the dog isn’t looking at you. We therefore train both. Dear old Gopher (cattle dog) was too smart for us, though. After she lost her hearing in her old age, if she was happily pottering around and caught a glimpse of a hand signal, she’d quickly turn her head…away. 🙂
Comments are closed.