More on Learning Theory
At Special Paws, we use operant and classical (Pavlovian-remember the salivating dog?) conditioning.
Operant conditioning involves pairing something the dog has control over (a behavior we would like performed for instance) with a consequence. For example,
your dog sits and he gets a reward. Here the dog is the operator on the environment: certain voluntary behaviors given by the dog will have certain consequences.
Behavior A simply results in consequence B. Another example of this is: your dog comes to you when you call him, so he gets a toy to play with.
There are 4 "quadrants" in operant conditioning:
In this sense, positive doesn't mean "good," it means the addition of something. "Negative" likewise means the removal of something.
Reinforcement is something that increases the incidence of behavior and punishment is something that decreases the incidence of a behavior.
Positive reinforcement is simply the addition of something to increase the incidence of a behavior. An example would be, you go to work and you get paid. Your
paycheck is the positive reinforcement for going to work (the behavior your boss would like to persist)! The addition of pay will increase my incidence of going to
work. I'm more likely to go to work in the future.
Negative reinforcement is taking something away to increase the incidence of behavior. An example is say I'm in a rush to make my flight and running through the
airport. Someone I'm traveling with grabs my arm and is tugging so we won't get separated. When I slow down, they let go. Because slowing down took away the
arm grabbing, I now have an increase the incidence of walking slowly. In theory, I'm more likely to walk slowly in the future to avoid someone tugging on my arm.
Positive punishment is the addition of something to decrease the incidence of a behavior. An example of this is if I reach for the last piece of chocolate cake and
someone smacks my hand away. The addition of the smack will reduce my incidence of reaching for the last piece of chocolate cake. In theory, I'm less likely to
reach for the last piece of chocolate cake in the future.
Negative punishment is taking something away to decrease the incidence of a behavior. An example would be if I start jumping up an down and flailing my arms
about to get someone's attention and instead they ignore me. They've now taken away the attention I sought to decrease the incidence of my jumping up and
down and flailing behavior. Now I'm less likely to do it again in the future.
Classical conditioning involves your dog's emotional response to a particular stimulus. For example, when your dog was very small, the sight of a leash meant
nothing to her because it was not associated with anything she knew. Over time, the leash began to predict her fun walks with you and an emotional response
grew, also known as a conditioned emotional response (CER). Your dog now associates her leash in a positive manner because it predicts the walk, so you see
a happy dance from her before each time you go for a walk. CER's come in positive or negative fashions since dogs view the world in one of two ways: as safe or
dangerous. For example, a dog who dislikes his harness may duck and dodge when you come carrying it to go for a walk. He associates the harness with
Counter-conditioning is a process used to change your dog's mind about something he already dislikes. This is most often paired with another process called
Desensitization. In that process, your dog is exposed to the thing that he dislikes at very small intensities, building up to full-blown exposure at some point down
the road. The dog mentioned above with the harness issue may be slowly introduced to the harness- first on the ground, getting a treat each time he looks at it
and gradually moving up to being able to hold it over his head and ultimately having him wear it. We are attempting to change his emotional response to the
harness while keeping him below the level of being nervous (also called "below threshold").
At Special Paws, we do not use negative reinforcement or positive punishment. These involve the use of aversives. Aversives by definition, imply the use of
something meant to frighten. In dog training this is through the suggestion that fear, pain or death could be a consequence of a behavior. These things often
include more than we could think, keep in mind that there are a number of behaviors that we might exhibit towards our dogs that could seem "harmless" enough
to us (such as leaning forward to shout at a dog), but have completely different meanings about consequences to our dogs.
Most of your training with Special Paws will involve positive reinforcement. You'll notice that in certain situations, we will use negative punishment as it is defined
here- all of which will involve the denial of attention as punishment. Under certain circumstances, we will employ counter-conditioning and desensitization.
(For excellent explanations and examples of all the above reference the books by Cristine Dahl and Jean Donaldson in our Resources section.)
We heartily encourage you to examine dog-human relationships and see how many of these facets of learning theory you can spot in daily interactions,
including formal training sessions.