One of my dogs used to sneeze whenever we’d walk out of the house to go for a walk. I thought that since he was so predictable, capturing and rewarding the sneeze would be simple. After about 10 trials, I noticed a positive CER* whenever I closed the door.
So, I thought I was rewarding a sneeze. Apparently, my timing was off and I was rewarding the closing of the door. About 15 more trials and he was finally sneezing on cue.
But that’s the beauty of reward based training. So what if my pup got excited about the closing of the door? My mistake was easy to fix. And my dog got treats so he was having fun either way.
When we use pain, discomfort, fear, startle or intimidation - mistakes are much harder to overcome – and of course it’s just plain stressful for the dog.
I’m walking my dog and he starts pulling because he sees another dog. I yank the leash and strangle my dog (euphemistically called a leash correction.) In my mind, I’m punishing him for pulling or I’m trying to “get his attention”.
But just because that’s what’s in my mind doesn’t meant that is what my dog is learning. A more likely scenario is that my dog starts to learn that other dogs equal pain. And eventually my dog starts becoming reactive towards other dogs. (negative CER)
I don’t want my dog jumping on counters so I set a booby trap. Something that makes a scary noise whenever his feet touch the counter. In my mind I’m showing him that touching the counter is wrong. But the dog might learn:
-trying to eat is scary,
-or going in the kitchen is scary,
-or he might just learn to be afraid of all popping noises in general.
-Or what if a neighbor kid walks by the window at the same time the booby trap goes off? Now the dog learns to be afraid of children.
Stay tuned for more on the problems with punishment
Follow us at https://www.facebook.com/groups/stubbypuddin/
*CER = Conditioned Emotional Response. In the positive example above, my dog started anticipating a treat whenever I closed the front door.