It’s unfortunate, but sometimes we train a cue only to find that it later becomes useless. If you train enough, this may have happened to you. But why does it happen, and what can we do to fix it?
In dog training sometimes we forget that the immediate consequence of a behavior is what causes that behavior to occur more or less often. We can, by accident, teach a dog to ignore something we have previously taught by pairing the cue with something the dog doesn’t like, or by taking the rewards away too quickly. Once this happens we have work to do in order to fix the problem.
I think of an example with a relative where we trained a very strong recall (come when called). The dog would recall even when playing with other dogs from across a distance the length of a football field. Yay! We had trained the recall and it was working! One day I was visiting, and I called the dog from about 20 feet and she ran the other direction. She would not come to me. The handler told me that her recall had fallen apart in the past two weeks. It was heartbreaking, because the recall had been a thing of beauty.
Upon discussion, I found out that the recall was only being used for one purpose: to call the dog from her romp through a twenty acre area of woods full of deer and squirrels and rabbits, and then to put the dog in her crate, usually when her human was going to leave the house for a few hours. This dog loved to run, did not like the crate much, and she really hated being left behind. So the cue came to mean that she would lose her freedom and her companion.
Had the handler used the cue a majority of the time for reasons the dog enjoyed rather than saving it solely for something the dog considered very unpleasant, we would have been ok. But now we are faced with a poisoned cue. And not coming when called is a huge deal, especially when you live where the dog can get very far from home and still be on the property.
Losing a cue does not mean we have lost the underlying behavior. A behavior taught with positive reinforcement will not just disappear, but we may have to go back and do some remedial work with our dogs. The second time around, we need to be very careful what the resulting reward or punishment is for the dog. Remember: we don’t get to define this. Our dog does. Some dogs like their crates. The dog in the example above did not particularly like it, and she really didn’t like being alone. So she was being punished for coming. Now we need to retrain her and avoid this scenario.
The first thing I’m going to do here is completely abandon using the cue that’s been poisoned. Unfortunately it was her name. I will talk to her and use her name when I praise her and love on her, but I will not use it for recall.
Next, I’m going to start from the beginning and retrain the behavior using a different cue. My choice here is actually a whistle because it’s loud and very different from the original cue. Because she knows the behavior, it’s going to progress faster than the initial training. I’m going to be sure to repeat the behavior several times a day with a really awesome reward and a big party every time she comes. This “practice” is not going to end once the behavior has been trained. We want to be sure that positive association sticks with her all her life.
There will be times we will need to call her for things she doesn’t like. That’s okay. Ideally, we will call her, give her lots of reinforcement for coming, and after about five or ten minutes, put her in her crate. The separation of the recall from the ultimate goal (for us) of having her safely in her crate must be enough that she doesn’t relate the two. If my relative is running late somewhere every once in a while and violates this rule it’s not going to set us back too much, but we need to be sure that the good results far outweigh the bad ones.
Another way we can practice a behavior when it’s in a situation where our dog isn’t going to want to listen is to reward them with the freedom to go back to whatever fantastic activity they were enjoying. For example, those times at the park we must always take care to call her from time to time, reward her, and then send her right back to her play. That way recall doesn’t mean the fun is over. At least not every time.
Recall is an easy cue to poison because we are trying to manage our dogs and get them under control, and they may not want that. But other cues can be poisoned as well. The best way to avoid it is to be sure that a cue that has been well trained is rewarded (results in something the dog likes) far more often than it is punished (results in something the dog finds unpleasant), and that it is never punished with something that results in pain or fear (physical punishment, shock, yelling at the dog, etc). If a cue hasn’t been well trained then nothing unpleasant should ever be the result of performing the behavior. Remember: reinforcement (good stuff) builds behavior, and punishment breaks it (and generally confuses the dog). In the scope of this post I am talking about behaviors that have already been trained.
Another example from my training past is a client who, against my recommendation but with complete compassion and understanding on my part for her desire to let her dog run, decided to have an invisible fence installed because a physical fence was impossible. Several behaviors we had taught began falling apart. I wasn’t sure what was going on until I showed up early one day and found the fence trainer* there working with the dog. I watched as the trainer used cues we had taught to lure the dog towards the fence (“let’s go” for leash work, “come” for recall, with a sit or two thrown in). The dog was finding the entire experience completely unpleasant because he was receiving shock from the collar. Our cues had disappeared because they had been associated with the impending shock from the fence. We had to go back and do some remedial work, but within a few sessions we had changed our cues, I had had a discussion with the fence trainer about his choice of words, and the dog was doing well, other than he didn’t enjoy being outside as much as he had when I first met him.
The bottom line is we need to be very cognitive of the consequences our dog faces for doing the things we have trained him/her to do. If the consequences are negative, the dog is going to “stop listening” to us. If the results are almost entirely positive the dog will continue to perform. And finally, if we screw it up, and we all make mistakes from time to time, we can and should go back and fix it so that we can enjoy our dogs more!
*I put this asterisk here because these folks are very well meaning, but their job is to install fences and teach dogs where they are, and they really don’t have an intimate knowledge of behavior. The gentleman I was dealing with was very nice and absolutely loved dogs, but had no idea why using our cues was a problem until I explained it to him.