Few days ago, I was watching Canis Film Festival nominee videos and saw something I've seen a few times in my training. In the video clip ‘Paco learns to close a drawer with his nose' at one point during the session, dog was spinning. I'm not sure how the spinning started since in the first scene, he is already spinning. Maybe the spin was shaped as ‘moving the head in the direction of the drawer' But, he was spinning and he received clicks very close to those spins. The rules change: he has to go near the drawer to get a click. But he is still spinning. And once, he, after a spin, gets near the drawer - and gets a click. Then, the dog does the same behavior: spinning and going near the drawer 4-5 times. Once, he just goes close to the drawer and gets a click. And then again, he spins and goes near the drawer. 4-5 times. At this point, I was pretty sure that dog thinks that spinning is an important part of what earns the reinforcement. He, though, fails to meet the new requirement: touching the drawer and time span between the spinning and actually touching the drawer gets much bigger and maybe here dog gets that drawer gets the clicks. Or maybe it was the word, I quite cannot distinguish which though. But I'm just guessing here.
Anyway, I saw this spinning and I clearly remembered the same spinning my dog did while I was shaping a relaxed lay down. I was so concentrating of getting the moment when he starts going towards the ground that I totally ignored the spinning. And during the session, after every spin he got closer and closer to what I wanted. He was on the right tract of laying down all the time. And since I didn't care about the spin, in the end of a session I had a dog that would every time spin and lay down nicely. When I realized it, I started shaping all over again, being careful not to shape anything else but the laying down part. I really don't know if I could get rid of the spinning part without doing it all over again because my dog was so sure he has to spin before he lies down.
So, as always I started thinking how careful we must be when we are doing shaping. Personally, I'm an shaping addict and I want to shape bunch of stuff but mostly I like doing shaping some interaction with an object just because I don't require him to generalize a lot. If I shape ‘bring the cookie-box', the cookie box is always on the same place. If I shape ‘throw into the bin' it's always the same bin. But I've also shaped my dogs to lay down just on her mat (so when I would say lay down she would go to her mat and lie down… a nice thing, but not what I wanted) or to circle me when she brings a toy. And I also thought few behavior chains like: bow and bark (instead of just bark) and beg and then say byby… I can only say that you do learn from your mistakes.
