I often hear from trainers that if I ignore a bad behavior, the dog will never do it again. It’s a very common advice that rarely works and is based on a huge misunderstanding of how rewards work.
The explanation is usually this one: animals will more likely repeat a behavior that is rewarded, and will stop doing something that remains unrewarded. Everyone knows that who ever trained a dog, right?
“My dog jumps on visitors.”
“Just ignore the dog.”
“It doesn’t work, I tried.”
“My dog is shaking under the table during fireworks.”
“Ignore her fear, and he will calm down. He will learn that he is not in danger.”
“But he seems to be more and more in panic.”
Well, well… In my opinion people who give such a useful and good advice (nah!), simply doesn’t understand learning theory and reward systems (or dogs).
Why does a behavior repeat even if it is unrewarded?
A reward, by definition, can be any stimulus, object, activity or situation that are naturally pleasurable, facilitate survival, or homestasis, or associated by these through learning.
The arrival of a guest is an exciting and joyful event. This affectionate and positive state is the reward of jumping and running around.
Sniffing, smelling something and hunting for it is a naturally pleasurable activity for dogs (that can also serve the survival of a hungry dog or the end of boredom while the owners are away), so digging in the garden can be highly rewarding.
For a fearful dog or a dog with territorial aggression, snapping will be rewarding, because the stranger will automatically back up, give space and the dog can finally calm down.
Barking at the fence is very-very rewarding. Dogs learn that if they bark, people will leave. In their smart heads, barking is associated with people leaving, and staying away from their property. People pass by on the street anyway, but they don’t understand that. Moreover, barking is naturally rewarding, so they will bark. Every. Single. Time.
Reward means not only a yummy treat, or a squeaky ball thrown by the owner, indeed! This is where the huge misunderstanding roots that leads to ineffective and potentially dangerous advice.
Rewards have three categories:
- Primary rewards facilitate survival. These are homeostatic and reproductive rewards. A homeostatic reward can be anything that results in a normal and calm state of the dog (mentally or physically), such as food, sleeping, snapping at strangers around a food bowl, or chewing on furniture. A good example of reproductive reward is the activity of licking bitch piss on the ground (Sorry.)
- Intrinsic rewards are unconditioned and naturally attractive and pleasurable like playing with other dogs. Being petted or simply being around others is also naturally rewarding for dogs since they live in families. Smelly treats are also rewarding even for a dog that is not hungry.
- Extrinsic rewards are motivating, because they gained their value through learning, associated with a primary or intrinsic reward, like praising and toys. Dog toys are not naturally valuable. They gain their value by playing with them. The words “good boy” don’t mean anything to a dog, but he learns that it is something pleasurable, because he is petted and given treats when he hears the words.
When a dog jumps at people, barks at the fence, snaps at strangers, steals food from the counter, runs away with other dogs to play, pisses the car, tears its bed into pieces are all very, very rewarding.
Where does this advice come from then?
It is based on the phenomenon of extinction. Extinction appears in classical and operant conditioning when a previously rewarded behavior does not predict reward any more, and the behavior likely disappears. The problem with this is that a huge majority of behaviors are primarily or intrinsically rewarding, and not something that was learned to be rewarded by treats and praise. As a consequence, the behavior will not disappear just because the owner ignores the dog.
Ignorance has bad consequences
You can look and walk away, behave naturally, or act as nothing is happening, but it will happen and will repeat every time until you do something about it. In a bad scenario, the behavior can escalate into an even worse behavior: snapping becomes biting, the cute puppy who jumps grows big, the panicking dog escapes – you hear the stories.
The only way to stop a ‘bad’ behavior (in commas since these behaviors are natural for dogs, it is us, humans who label them as bad ones) is by understanding the mechanism how dogs learn, by discovering the individual differences a particular dog has regarding rewards, and by positive training.
Only understanding and learning can result in stopping unwanted behavior for dogs – just as for their owners and trainers.
