What if your dog’s cue for a behavior is not what you think it is? Can you be sure–absolutely sure–that the dog really understands what you want?
That’s another place where punishment based training can really go awry. How often are dogs punished for failing to perform when they just don’t understand? I think it’s much more often than most people realize.
My dogs are pretty good at sitting. But their cue for sit and stay is not what I intended it to be. Luckily I figured that out. I know exactly where the problem came from and how to fix it, should I decide to. And perhaps by sharing, I’ll help some others who might be in the same boat.
The Three D’s
Dog trainers often talk about the “three D’s” in training: duration, distance, and distraction. Each of these represents a set of challenges for the dog.
When working up a stationary behavior such as a sit/stay (or even a moving behavior such as walking on leash, but let’s limit the discussion to stationary behaviors for now), you need to gradually work up the length of time the dog can do it. That’s what we call duration.
But even if your dog can hold a sit/stay for 60 seconds while you are standing right there that doesn’t automatically mean she can do it if you park yourself 10 feet away. That’s distance. Distance needs to be specifically and gradually taught. It’s a challenge for several reasons. Principal among them is that for most of their early training, dogs learn that the reinforcement usually happens right there on your person or close by. They will tend to follow you when you move away.
And even if your dog can hold her stay while you quietly stand 10 feet away from her there is still another aspect of challenge. That’s distraction. Can she stay if you drop your treat bag? If another dog in the family trots by? If you go sit in a chair (surprisingly hard, since dogs often associate that with the session being over)? If the doorbell rings? If her best friend comes in the room? If you toss her favorite toy in her direction? You get the picture.
Most trainers train the three D’s in roughly the order above. Get a little duration first before introducing the other challenges. Then train distance, then distraction. (Distance is just a particular kind of distraction, anyway.) But what happens if you jump straight to the other challenges early on? You’ll see in the movie.
My Achilles Heel: Duration!
When teaching the very beginnings of duration, one usually silently counts seconds. There are different protocols for going about this, but the idea is to gradually lengthen the amount of time the dog stays in position without being released. With behaviors like holding a dumbbell, you usually have to start with increments of 10ths of a second to get that first bit of duration. With sits and downs, the dog can generally already perform the behavior for a second, maybe two, so you start there and work up.
I get terribly bored just standing there teaching duration, so my most common error as a trainer is starting to move around too soon. What has happened as a result is that my moving away has become part of the cue for “stay.” Oh-oh. If I don’t move, they don’t believe me. You can see that in the video.
Here’s how long each dog lasts on her sit/stay when I just stand in front of her.
- Clara: 4 seconds until she breaks position to nudge my hand, and 4 more until she stands up.
- Summer: 1 second until she goes into a down.
- Zani: 1 second until she goes up into “sit pretty,” another 2 seconds until she goes into a down.
That’s pretty embarrassing. But also note in the video that when I move away immediately after giving the cue, they can hold their sits successfully through duration, distance, and distractions. They can all last several minutes when I do that, with distractions including tug toys dropped in their vicinity, my walking around or actually leaving the room, and all manner of food placed out to tempt them. Again, their cue to stay is my movement much more than my verbal cue.
Edit, 2/9/16: In the movie, I dubbed in the word “Fail” when each dog broke her stay, to mark how fast that happened. A viewer’s comment caused me to realize it sounds like I’m using a No Reward Marker. I wasn’t. It’s only in the voiceover, not from the training session. And in case it’s not clear: the failure is never theirs. It’s mine.
What Did They Do Instead of Staying in a Sit?
Each dog switched to a behavior that had been recently and heavily reinforced. Summer went into a down on the rug. She had a tactile cue (soft surface under the butt) to perform her mat behavior. Zani tried “sit pretty,” which we have been working on in that very place, before trying a down as well. Clara tried a stand, which we have also been working on in that area of the house.
It’s a beautiful example of the dogs’ reinforcement history taking over when they are unsure. I didn’t give what to them has become the “real” cue for a sit/stay: for me to move away and start doing random things and providing distractions. So each dog gave up but each tried something different: what has most recently been reinforced in that context.
This kind of thing happens all the time. Frequently we don’t actually know what the cue is for the dog. And in the absence of that special cue, they revert to guessing and expressing their reinforcement history. It is often a reason people think their dogs are “giving them the paw.” That’s very sad, since what the dog is doing is trying to get it right in the face of unclear instruction.
Besides the fact that the cue for my dogs to stay in a sit is for me to put them in sit and immediately leave (and I didn’t do it), there are a handful of other reasons for their failure to sit. As I mentioned, reinforcement history contributes. A lack of confidence in verbal cues in general (because I don’t always work hard enough on cue recognition and stimulus control with them) is another. The fact that I generally encourage them to offer behaviors is yet another; they have nothing to lose from guessing. And finally, though I hate to mention this, I don’t use a “stay” cue. I’m not a good example of that practice, which can work perfectly well when one is clear about the cue to begin with! But adding in a “stay” could probably make up for some of my other frailties as a trainer.
Can They Really Not Do It?
Of course they can hold a sit/stay with me standing right there. I’ve mentioned before that I do a lot of my training following Sue Ailsby’s Training Levels. One of the early steps of sit/stay (Level 2 Sit) is to work up to 60 seconds when you are standing in front of the dog. All of my dogs have performed that stay many times, starting early on in our training relationship. But in the course of life, I don’t use that behavior much. I don’t ask my dogs to sit for long periods, especially with me standing right there. I use down much more often, so the sit/stay breaks down.
But it takes only a few minutes to get that sit/stay duration back if I want it. Even in the course of filming the above “failures,” the dogs started working up their duration again. I could have a minute long sit/stay from every one of them after a couple of training sessions. But honestly, I’ll probably let it lapse again.
In case you would like to see a lovely “real life” sit/stay, here is Zani doing a short agility run (in which we won first place, I might add). Being parked in front of a jump standard is another clear cue for her to hold a sit/stay.
The Map of Reinforcement
I’ve mentioned my concept of the “map of reinforcement,” something I started to see once I learned and really internalized that behavior was driven by consequences. I wrote about this at length in my post, “What Dog Training Really Taught Me.” What the dogs tried instead of sitting and staying showed a map of what had been reinforced. And here’s something related: my dogs’ behavior is also map of my behavior. Yep. Whenever all three of my dogs have the same “failure,” it’s a pretty sure bet that it is something that I am consistently doing. An experienced trainer could look at the movie and know exactly what I did when training to cause my dogs to break their stays as they did. I know what I did, too. The fact that I don’t always fix stuff like that is perhaps why I’m a writer and not a dog trainer!
Care to share situations where you found out that the cue for the dog was not the cue you thought you were giving? And has anybody else done this particular silliness with duration behaviors?
- Dogs Notice Everything–some examples of why dogs would fail to respond correctly to a cue
- Missed Cue Video Playlist–5 videos that show dogs failing to respond to cues for perfectly explainable reasons
- 16 Behavioral Cues that I Didn’t Train (But are Still for Real)–emphasizes the important of non-verbal, non-deliberate cues
- The Right Word: Reducing Errors in Verbal Cue Discrimination–teaching my dogs separate release cues using a reduced error method
- Good Sit!–why it’s not the greatest idea to say, “Good sit!” after a dog performs a sit
Copyright 2015 Eileen Anderson