My last blog explored the phrase “It depends on the dog”, and discussed this concept within the context of problem solving.
Today we’ll consider the same topic but I’d like to discuss how one might approach teaching a new behavior (as opposed to problem solving) with different dogs. To illustrate this, I’ll use the three dogs that live in my house as examples. Each dog is going to learn the same behavior and none have any experience with it. What they do have is different temperaments and training histories, which in turn drive my decisions. Let’s look.
The behavior that I want to teach is a simple jumping exercise. In a nutshell, I want to teach each dog to jump back and forth while I kneel on the side of a jump with a treat held low near the base of the jump.
Dog #1: Lyra. Lyra is a great dog for shaping. She has a calm head and doesn’t get stuck or frantic with behaviors. She is quick to identify what causes the reward and eliminates incorrect options with minimal stress. In Lyra’s case, I started with the final behavior except that I set the bar somewhat low (approximately 8″), knelt by the side of the jump and simply waited. Because she has a background with jumps and jumping, she offered a jump within a short period of time. At the end of a few minutes she was relatively proficient at the exercise. Within three sessions, she had eliminated all of the alternatives (down, backing up, pawing at the jump) and was consistently offering only the correct behavior. She took the easy route for both of us!
This is what it looks like with Lyra after a few days – we pretty much started the same way it looks here:
Dog #2: Brito. Brito is also a good dog for shaping, but if he gets frustrated he’s quick to leave the lesson, and since he finds alternative interests very quickly, it was important to me that he “find” the correct answer quickly. In Brito’s case, I started with no bar at all on the jump. I knelt by the side of the jump, but I also rocked my body back and forth to encourage him to cross the ground bar and also to cause mild engagement based on my movement. I was also close enough to the jump that coming between me and the upright was not a likely choice. Once he was crossing back and forth easily, I added the bar and raised it to his current height of 6″ for this exercise. Approached in this manner, Brito also learned the final behavior rather painlessly within a few sessions.
This is what Brito’s process looked like. As you can see, if I stop engaging him for more than a second or two he finds other things to do, so I had to keep him busy (some days I really wish I had an indoor training facility!):
Dog #3: Raika. Raika is an older dog with a good deal of lure based training in her background. Almost all of her shaping was done more than eight years ago, and most of it was structured shaping (errorless learning) as opposed to free shaping (more guessing with less guidance).
At first I tried free shaping this exercise and Raika offered a huge range of options. These included standing on the bar, down, sit, back up, bark, bite the bar, bite the jump, drag the jump several inches, use her nose to shovel underneath the jump (knocking the entire thing over), crawling under the 6″ bar, etc. After two minutes of that, she left me altogether and started making frustration based mischief in the training yard. Because I prefer to avoid excessive frustration in training, that was a pretty clear signal to try something else. Enough is enough.
Raika knows how to jump on cue, so my first “fix” was to simply cue her to jump. She didn’t understand – I’ve never kneeled on the side and asked her to jump. Next I tried standing next to the jump and cueing a jump but by this time she had offered so many behaviors in the prior session that she was shutting down. So I started over.
I placed Raika away from the jump in “broad jump” position and cued a jump, simultaneously throwing a cookie forwards. She understood what I wanted and asking for this familiar behavior “unstuck” her. From there I tossed the cookies closer and closer to where I wanted her to start landing (next to the jump) and I started lowering my body posture to the kneeling position. The last thing I added was holding the treat in my hand down low by the upright. Here is the progression that made sense for Raika:
Raika’s process took twice as long as that of the first two dogs because rather than figuring it out for herself, I had to provide a good deal of direction. That’s ok. There is no race, and at the end of the day all three dogs are able to perform.
The moral of this blog post….train the dog in front of you. If you are kind, take your mutual training history and your dog’s temperament into account, and change course when you realize you are going in the wrong direction, then you’ll get where you want to go.