Alexa Capra | Dogs and More - The positive reinforcement fallacy and its fallouts

For owners, and sadly also for many trainers, positive reinforcement refers to the consequence of a behavior.

If your dog is displaying a “good” behavior, the consequence has to be positive. Dogs are supposed to get good things in life if they behave well, and be ignored, or experience negative consequences, if they misbehave.

This interpretation of the positive reinforcement principle has so many negative effects on dogs, humans, and their communication and relationship.

I receive a phone call, many years ago. A friend of mine is telling me that her dogs had a fight. She lives in a small flat with 5 dogs, some of them with physical disabilities. What will you do, about it? I ask. Her answer has a huge impact on my life: “I will wake up earlier, spend more time cuddling every one of them, and more time walking them in the woods. I will work harder to make them happy.”

german shepherd fench bouldogue

Kandji and Sonne

Three days ago, I am thinking of those wise words, as I am dealing with a situation at home. One month ago a french bulldog puppy joined our family. I live with three german shepherds and a border collie. Arj was motherly protective from the first sight. Grisou was simply pretending she wasn’t there. Puma decided she should die. Kandji was scared of her, and aggressive every time she tried to approach him.

One of my dogs could harm her or even kill her with one bite. I was using Liviana’s strategy to make my dogs accept Sonne: I’ve walked every day in the woods with Sonne and one or two of my dogs. It worked pretty well for all of them. They all decided she could live.

I was able to relax a little, and go back to our normal life. This means Kandji was allowed to come upstair, in my bed room, and have his share of cuddles on my bed, in the evening. The first evening he couldn’t stand her, and went downstairs to avoid her. The second evening he was on my bed, getting stiff and growling as she tried to jump on it. The third evening it was even worse, he almost attacked her. It wasn’t working. I got scared, I stopped Kandji and told him to get out of my room. No dog is allowed to harm a puppy, on my watch. He was so stressed.

During the next day I was unhappy about what happened, and planning a better strategy. I decided to go back to Liviana’s principle. That evening he came to my room, and I focused only on him, telling him how good he is, how much I appreciate him, that he doesn’t need to worry about Sonne. He was able to relax. In the next two days it went better and better. In three days he was able to tolerate her on the bed, not getting stiff, not growling, not threatening her even when she approached him.

puppy dog on bed

I took a picture of the first time Kandji was able to tolerate Sonne without getting stiff and growling/threatening her

Remember, the positive reinforcement? That night his previous “bad” behavior had the most positive consequences possibile. Yet, he improved.

Of course, the positive consequences were not associated with the growl and the threats, and I should not refer to positive reinforcement in this case (there is no direct association between the behavior and the consequences). I am talking about how our mind process and applies the principle of positive reinforcement. We wait for the good behavior, we ask the dog to behave well, we want and pretend the dog to behave the way we like. Only then we produce a positive consequence, believing this is the right way to educate a dog. A dog only deserves something good if he/her behaves the way we want.

This is the fallacy and the fallout of the positive reinforcement principle.

We focus on the behavior-consequence association, and completely forgot about the causes, and the outcome of the experience. The reinforcement principle is NOT about positive or negative consequences. It is about the future. A consequence is a reinforcer only if it increases that behavior in the future. You can only assess the efficacy of a consequence through the future choices of the dog.

1. Stop focusing only on good/bad behaviors. Learn to think in terms of emotions and motivations.

2. Stop ignoring the “bad” behaviors. Watch, understand, communicate, help. Do NOT ignore your dog. It’s so fucking rude.

3. Stop reinforcing the “good” behaviors to get what you want. Be happy, be proud, communicate with you dog, show him/her that you are aware and you care. Be honest and be real.

4. Do not wait for a “good” behavior to give something positive to your dog. Make your dog happy, safe, and good behaviors will be a consequence of his mental and emotional state.

5. “Bad” behaviors can improve through positive consequences.

6. “Good” behaviors do not always increase in frequency through positive consequences.

7. Stop trying so hard to change your dog’s behavior. Focus on understanding the causes and motivations, give your dog better tools outside the problem.

8. Stop using the scientific principles of learning as you were cooking a cake. I want more of this, less of that. It can’t work. Learn to know your dog, be aware of your own emotions and motivations.

9. Food is a positive consequence if the dog is feeling safe and hungry. A treat is not a positive consequence if the dog is scared of something someone. Feeling safe is a positive consequence.

10. You are not a bad owner if you are unable to change a “bad” behavior while your dog is displaying it. The dog might be unable to listen, unable to make a different choice, in that moment. We are unable to control our emotions and our behavior when in an intense emotional state. How possibly can we have the power to control someone else’s emotions.

The learning principles, and the motivation to display a behavior.

You might think that the behavior-consequence is a strict rule in a learning setting. Well, no. Your dog has learnt to touch a target mat with the front feet. You ask your dog to run to the target. The dog doesn’t. What will you do? Ignore the dog until the dog touches the target? Yell at the dog? Help the dog reach the target? For sure, you will not give food or a toy to the dog, the dog only deserves a reinforcer for the correct behavior. Try. Try to reach the target, help the dog complete the task AND give him a really positive reinforcement. Tell your dog he has been great, you’re so happy he was able to touch it, and give him a lot of food or play a lot.

Stop focusing on behavior-consequences.

Think in terms of causes and motivation (future choices).

If you help the dog and you reinforce the behavior, you are making the dog successful, and you are building a motivation to display that behavior in the future.

Do not reinforce to get what you want. Be happy, be real