The Four Quadrants of Dog Training
Learning Theory is a Good Investment
An Important Note
In behavior terms, the words positive and negative are not used to denote good or bad. Positive simply means addition and negative simply means subtraction. Also, in behavior terms the word punishment, is meant to denote a behavior that decreases in frequency, whereas, reinforcement denotes a behavior that increases in frequency.
What are the Four Quadrants of Dog Training?
If you have never heard about the four quadrants of dog training, no need to worry; most likely you are using them all the time when you train your dog or work on behavior modification. You don't need to be a certified dog trainer or train dolphins at sea-world to learn the dynamics; all it takes is just understanding how dogs react to their surroundings and how their environment affects them. In this hub, I will try my best to explain these quadrants in an easy-to-understand manner. Most of the books I have use quite technical lingo and things seem to get confusing and murky when they don't need to be.
What are the four quadrants of dog training? They are four outcomes, or better consequences, that occur when your dog interacts with you and the world around him. A dog reacts accordingly to these quadrants, depending on exactly what is happening.
Dogs learn through operant and classical conditioning, The term "operant conditioning" takes place when a dog associates a behavior with a consequence. According to Thorndike's Law of Effect "responses that produce a satisfying effect in a particular situation become more likely to occur again in that situation, and responses that produce a discomforting effect become less likely to occur again in that situation."
These consequences take place internally and externally. Externally, such as when you provide a consequence for barking by telling your dog to shush, internally such as when your dog feels hot and he seeks a shadow under a tree. It's important to note that consequences you perceive as pleasant or unpleasant don't necessarily have to be the perceived the same way by your dog. How consequences are perceived are highly subjective and may vary from one dog an another.
For instance, some dogs stop barking and feel intimidated by owners who reprimand them; whereas others who crave attention because they have been alone all day, may enjoy the extra attention even if negative and will continue to bark. Another example? the heat from the sun may feel terrible from a Saint Bernard's point of view, whereas, the warmth of the sun may feel wonderful for a Chihuahua. For more on this and a story about a Lab I trained to stop persistently barking read my article: Dog Training The Use of Rewards and Punishment
Ready to learn more about the four quadrants of dog training? It is astonishing, the amount of websites, some even written by so-called "professionals", who continue to promulgate confusion among the terms negative reinforcement, negative punishment and positive punishment. We'll take a look at them one by one and hopefully this should help clarify. I will provide some examples for each quadrant for clarity sake. So here they are: positive reinforcement, negative reinforcement, positive punishment, and negative punishment.
Positive Reinforcement explained by Victoria Stillwell
This is my favorite method of training. I love it because it's gentle, effective and as Thorndike mentions: "Responses that produce a satisfying effect in a particular situation become more likely to occur again in that situation". As opportunistic beings, dogs indeed tend to repeat behaviors that have a satisfying effect. Have you ever wondered why dogs beg at the table? Because they're rewarded with food. Why do dogs jump on people? Because they are showered with attention. Why do dogs counter-surf? Because they find little treasures.
What's the "law" of positive reinforcement? In positive reinforcement, the term "positive" means adding something for the purpose of making a behavior increase (reinforcement ) So if you give a dog a cookie every time she sits, you will see her sitting more often, especially if she loves cookies! If your dog felt good sleeping on the sofa, he'll likely jump on the sofa more often. If a dog jumps on his owner, and the owner pushes him away but the dog likes the attention because he has craved every bit of it all day long, he'll likely jump more.
Tools meant to deliver positive reinforcement include clickers, target sticks, treat dispensing machines, tug toys and anything your dog loves. Methods meant to reinforce behaviors include praising, petting, giving treats, playing with a toy or any life reward that your dog perceives as as good. For more on life rewards, read my article on the Premack Principle.
An Example of Negative Reinforcement versus Positive
I'm not a fan of negative reinforcement, simply because it often involves things dogs perceive as bad. Yet, it's also true that at times, you can use negative reinforcement in a minimally aversive way such as in BAT training. What are the "laws" of negative reinforcement? In negative reinforcement, the term "negative" means removing something for the purpose of making a behavior increase (reinforcement). In this case, we're looking at removing (negative) something for the purposes of increasing behavior (reinforcement). This sounds a tad bit hard to understand but it'll make sense once you look at some examples.
For instance, imagine again your dog is under the sun. The dog doesn't like it cause it's burning his skin and making him feel very hot, since the sun is perceived as bad, the dog will want to move away to feel better. So he'll go under the shady tree for relief. Because being in the shady tree feels good, next time when he's slightly hot again, he'll likely keep on going under the tree when he's hot (reinforcement).
If you ever worked on training your dog to sit you may have applied negative reinforcement without knowing it. Say your dog wasn't sitting and you decided to hold him by the collar to guide him down or push his rump. Most dogs don't like the pressure, so they are reinforced when they sit and you remove that sort of pressure (negative). Soon, the dog will sit more and more (reinforcement) just to avoid that pressure. (I am not fond of this method, nor do I recommend it). If you like to go horse-back riding, you'll use negative reinforcement. The horse knows that you remove the pressure on the reign once he turns or stops. He also knows that you stop putting pressure on his side with your boots once he starts speeding.
Shock collar trainers use negative reinforcement a lot. They like to deliver continuous shock until the dog does the wanted behavior. For instance, when training a recall, they will call the dog and deliver shock and release it only until the dog comes running to them. These brutal training methods cause unnecessary stress in dogs and many trainers won't even take the time to train the dog, what to do in order to escape the shock.
Presentation on why Positive Punishment Shouldn't be Used
I'm really not a fan of positive punishment, simply because it involves again things often perceived as bad by dogs. In this case we're looking at adding (positive) something in an effort of stopping an unwanted behavior (punishment). Remember Thorndike's Law of Effect "responses that produce a discomforting effect become less likely to occur again in that situation." Let's look at some examples.
For instance, imagine your dog sees a skunk and he goes to pester it. He then gets sprayed, something he perceives as bad, and as a consequence, he never pesters a skunk again (punishment). Of course, there are many dogs who will continue to pester skunks no matter what, which simply confirms that not all dogs perceive things in the same way, indeed, for some dogs the adrenaline of the chase outweighs the risk of being sprayed.
Need another example? You puppy pees on the carpet and you use a rolled newspaper to punish him. The addition of the newspaper (positive) makes your puppy reduce his behavior of peeing on the carpet (punishment)
The "law" of positive punishment means you are simply adding something to make a behavior decrease.
Products meant to deliver positive punishment remotely or by the dog owner include scat mats, bark collars, electronic fences, shock collars, choke collar, prong collars , rolled newspapers and any other tools meant to be perceived negatively by a dog. Keep in mind that misused, even a buckle collar and a leash can be used as positive punishment. Methods based on positive punishment include collar grabs, alpha rolls and other aversion-based methods meant to intimidate dogs.
An Example of Negative Punishment
Despite the word "negative" I like to use negative punishment as dogs seem to respond to it and it's an effective way to stop an unwanted behavior without resorting to stimuli the dog perceives as bad. In this case, we are removing something. Let's take a look at some examples.
Your dog jumps on you, because he's happy to see you come home. Since your dog is doing this for attention, the best thing to do is to turn around and become boring. By turning around you are removing something the dog perceives as good (the attention), for the purpose of eliminating a behavior. (punishment)
Need another example? You are training your dog to perform a behavior but he fails to perform it. In this case, remove the treat you were about to give him and give it to your other dog or eat it yourself. In this case, you remove something your dog likes for the purpose of extinguishing a behavior.
The "law" of negative punishment means you are removing something to make a behavior decrease.
Methods meant to use negative punishment include, time-outs, removing attention, removing treats, withholding the click of the clicker. This is often accompanied by a negative marker.