Learning by Consequences
You’ve probably heard the term “positive reinforcement” when it comes to dog training and visions of pockets stuffed with liver treats dance in your head. But what is “positive reinforcement” really?
Positive reinforcement is one of the four principles in operant conditioning and simply means that if something pleasant follows a behavior, that behavior has been reinforced and will get stronger. Operant Conditioning is so called because the animal (dog or human) “operates on” or controls the environment by choosing to behave based on what they expect the consequences to be.
Here are all four of the principles:
Positive Reinforcement (R+)
Negative Reinforcement (R-)
Positive Punishment (P+)
Negative Punishment (P-)
To understand this better, you need to know that the words “positive” and “negative” have nothing whatsoever to do with “good” or “bad”. “Positive” means something has been added (maybe something pleasant, maybe something not so pleasant). “Negative” means something has been subtracted, or removed (maybe something pleasant, maybe something unpleasant).
As previously mentioned, “reinforcement” means something has been strengthened. “Punishment” means something has been suppressed. In operant conditioning, that “something” is a behavior that the animal chose to do. Regardless of which principle is being applied, the beauty of operant conditioning is that the animal chooses to behave in a particular way based on what experience tells them the consequence, or outcome, of that behavior will be.
Here are examples of operant conditioning in our human environments:
Positive Reinforcement
You study hard for an exam and get an A+. The behavior of studying has been reinforced (strengthened) by getting a high grade (something pleasant has been added). You are more likely to study hard for the next exam.
Negative Reinforcement
You put on your seatbelt and the annoying buzzer stops. The behavior of putting on your seatbelt has been reinforced (strengthened) by the removal of the annoying buzzer sound (something unpleasant has been removed). You are more likely to put on your seatbelt the next time you get in the car in order to avoid that annoying sound.
Positive Punishment
You’re late for an appointment and run a red light. A cop gives you a ticket and a hefty fine. The behavior of running a red light has been positively punished (suppressed) by getting a ticket and a fine (something unpleasant has been added). You are less likely to run a red light in the future.
Negative Punishment
Employees get a year-end bonus for perfect attendance. You have never missed a day’s work this year, but at bonus-time your boss tells you money is tight and there will be no bonus. The behavior of never missing work has been negatively punished (suppressed) by not getting the expected bonus (something pleasant has been removed). You are less likely to strive for perfect attendance next year.
So how does this work in dog training?
Positive Reinforcement (R+)
The behavior causes something good to be added into the picture:
The dog sits. A treat is delivered.
The behavior of sitting has been positively reinforced and is likely to be repeated.
Negative Reinforcement (R-)
The behavior causes something unpleasant to go away/be subtracted from the picture.
The dog opens his mouth when you present a dumbbell and pinch his ear. The pinching stops.
The behavior of opening the mouth when the dumbbell is presented has been negatively reinforced and is likely to be repeated.
Positive Punishment (P+)
The behavior causes something unpleasant to be added to the picture.
The dog chases the cat. You deliver a shock from the e-collar.
The behavior of chasing has been positively punished and is likely to decrease.
Negative Punishment (P-)
The behavior causes something pleasant to be taken away.
The dog jumps at you to make you throw the ball. You put the ball away.
The behavior of jumping to make you throw a ball has been negatively punished and is likely to decrease.
It is important to note that there is likely to be fallout from the use of positive punishment and negative reinforcement and that both, if you choose to use them, must be implemented wisely. Remember that animals (dogs and humans alike) seek to avoid unpleasantness. Dogs are expert at avoidance, but they don’t necessarily have to leave the situation to avoid the unpleasantness. You might see a dog who is being trained with aversive methods “shut down”, becoming apathetic and refusing to do anything at all; it is safer for him to to nothing than to risk the unpleasant consequence of a mistake. Some dogs will use aggression as their means of escape/avoidance, and the aggression may be directed towards any person, animal or thing nearby. (I once witnessed two Dobermans running towards their invisible fence barrier. When the first hit the fence and felt the correction, he turned and attacked the dog behind him).
Most trainers today prefer to use positive reinforcement, and sometimes negative punishment, as the main tools in their toolbox since there is little evidence of any fallout from these methods. Reinforcement, however, does not always have to be a food treat. Whatever you use to reinforce a behavior has to be something the dog really wants at that time. In certain situations, that may not be food! It may be chasing a ball, or getting to go outside, or going for a car ride.
Years ago I had a Beagle mix who should have been named ‘Houdini’ for his ability to escape our yard. He would always come racing back when called because as soon as he returned he would be met with an open car door and a ride around the block. Car rides were one of his favorite things in life, so the car ride reinforced his choice to come back. If I had yelled at him, the yelling would have been a positive punisher for coming when called (not for running away) and he would have been less likely to return the next time.
Disclaimer: Allowing dogs to ride unrestrained in vehicles in not condoned by the writer.
Understanding how dogs learn by the consequences of their choices will help you, no matter what you are trying to train (or un-train) your dog to do. Try to set up training scenarios so that you can positively reinforce the dog's choice. The consequence for you will be a happier relationship with your dog!
