Train Your Dog the Psychological Way
Teach your dog all the basic commands with a few principles of learning theory.
Posted Apr 09, 2014
With baby boomers reaching retirement and empty nests, there is a resurgence of dog ownership. Dogs make wonderful companions. They are loyal, patient, loving, and non-judgmental. They warn us when someone is coming, they welcome us home, and they get us outside and moving. Dogs are good for us. But to live harmoniously, we need to train them and communicate with them.
Interacting with dogs involves both sending and reading communication. And, communication is both verbal and non-verbal. While some dogs do learn words, communication (even human communication) is primarily non-verbal. With dogs this means primarily voice tone and gestures. With humans, facial expressions, eye contact, and body movements are very important, too.
Dog trainers focus on the sending side of communication. So we will start here. But, reading dogs is just as important and I will take that up later. The absolutely mind boggling thing is that humans and dogs can communicate at all. After all, this is cross species communication. Misunderstanding between humans is common place so that dogs can understand and follow our directions really amazes me. Anyway, dog trainers refer to sending communication or giving directions, commands. Basic commands include: sit, down, stay, heel, no, out, come, and of course the dog needs to learn their name. Teaching these commands relies on universal laws of learning theory. Trainers are often unaware of it or use the terminology incorrectly but what they are actually doing is applying principles of classical and operant condition.
Just for the sake of review, remember Pavlov conditioned dogs to salivate to a light. A light was turned on (CS, conditioned stimulus), dogs were given meat powder (US, unconditioned stimulus), and they salivated (UR, conditioned response). The conditioned stimulus is followed by the unconditioned stimulus , CS → US and this is reinforcement. After several reinforcements the dog salivates when the light is turned on. Skinner took classical conditioning one step further and called it operant conditioning. Where classical conditioning relies on a response that is automatically tied to the stimulus, like meat powder and salivation, operant conditioning teaches a novel behavior. Here is a simple example of conditioning. Let’s say you want to teach your dog to sit. One way to do this is to start with your dog standing, facing you and with a closed fist with doggie treats. Lure the dog toward with the treats, pull your closed fist up, back and over the dog’s nose. The dog will be forced to sit down. Praise the dog with “yes” or “good” and give her a treat. Repeat this several times then pair the action of sitting with the word, “sit.” Gradually, ease off the treats. With repeated trials, the dog will sit to your verbal command and you can ease off the treats, just as Pavlov trained dogs to salivate to lights without meat powder and Skinner taught pigeons to peck keys without food pellets. For a very nice example of learning theory applied to dog training T.F.H. Publications has a series of books on several breeds of dogs that come with training videos. My copy is titled, The Toy and Miniature Poodle by Janice Biniok but they publish very good books for nearly every breed of dog.
Now, this leads to other topics like reinforcement schedules, shaping, and generalization. The question that came up in a recent dog obedience class was “do I give a treat every time the dog sits or do I use a partial reinforcement schedule?” The teacher was busy and distracted and I don’t think she knew about reinforcement schedules but she shouted out over the barking dogs, “every time.” What Skinner and other behaviorists learned was that positive reinforcement was necessary a 100 percent of the time when an animal was first learning something. No reinforcement will result in extinction or the animal stopping to sit, peck keys, or run a maze and intermittent reinforcement is the strongest reinforcement schedule. So a better answer would have been, “yes at first, when the dog is first learning to sit or come, give a treat every time.” “Eventually, give a treat every third time, then every tenth time.” Also, after the dog has learned the desired behavior, they will be able to do it for your praise and affection and dog treats may not be necessary. “If the dog stops performing (an extinction schedule), start over with the 100 percent reinforcement schedule.”
Shaping, successive approximations, and generalization are important principles of learning theory, too. You probably won’t get your dog to lie down on the first try but you can get something close. If you train the dog to sit, then put the treat down in front of her, and pat the ground, she will lie down. Give the treat, praise, and pair the behavior with the word “down.” Shaping is the concept of starting with an approximation of the target behavior and then getting closer and closer. Each successive approximation is closer to the target behavior. Some people will recognize this behavioral principle as “baby steps.” Teach your dog in successive approximations or baby steps. Reward each approximation until you reach the desired behavior.
Finally, generalization. It is not enough to teach your dog to sit on the comforter in the corner. I want her to sit on command, at the street corner, at the school, and when friends come over. This is referred to as generalization. This is harder. The dog may be distracted. But, with practice of these same principles: operant condition, reinforcement, your dog will learn to generalize sitting behavior to many settings. In sum, if you can master, apply and practice a few general principles of learning theory such as operant conditioning, reinforcement, shaping, and generalization you will be able to teach your dog all of the basic commands. You and your dog will live in peace and harmony and enjoy many fun and loving years together. Sound easy? OK, now go try it and get back to me.