Most households with a dog are familiar with it, the doorbell rings and the dog starts barking. “I never taught him this, but he does it anyway” is the response of many owners. Also jumping up against the visitors, something we never consciously taught him, did we? In this article you can read more about the way dogs learn and how we can influence this.
Many households in the Netherlands have a dog, a nice companion to play with, to walk with and often an effective way to deter burglars. In general, it is very nice when the dog listens and most people go with the dog on a course, some only go to puppy training, others really like it and also train or exercise with the dog. There is also another group of dog owners who enjoy working with the dogs, these working dogs are often trained on security work or police dog training.
On puppy course we are among very cute puppies. With the help of a biscuit the dog is lured into a sitting position and after a while the dog can sit on command. During training sessions of 10 years ago or with working dogs, people often have the idea that the “sit” is taught very differently, there is pushed on the back until the dog is seated. I don’t want to talk about whether this harms animal welfare. The fact remains that both methods work, but they are based on a different learning principle.
Classical conditioning
Classical conditioning is a form of association learning in which a dog starts to make the connection between two stimuli. Classical conditioning was discovered accidentally in 1905 by the Russian scholar Pavlov. Pavlov studied the functioning of the salivary glands in dogs and fed the dogs food to induce drooling. After a while the dogs started to drool when they saw his test setup, while there was no food at all around. The dogs had learned to make a connection between the test setup and the food.
In the same way, dogs have learned to bark when the doorbell rings, most owners have never been aware of this. The dogs bark when there are visitors at the door. After a while, the dogs start to see the connection between the doorbell ringing and the presence of visitors, so they start barking when the doorbell rings. It can go even further; perhaps the visit will walk on a gravel path in the front yard before they ring the bell and are at the door, after a while the dogs will learn the connection between walking the gravel path and the doorbell, with the result that the dog will bark when someone is walking on the gravel path.
Teaching the use of the clicker also falls under the classic conditioning. When teaching this, a click is made a number of times followed by giving a candy. The dog will soon begin to see the connection between the click and receiving a cookie.
Operant conditioning
Operant conditioning is a form of association learning in which a dog starts to make the connection between its behavior and its direct consequences. If the consequences are pleasant, the dog will start to display this behavior more often. If the consequences are not pleasant, that behavior will diminish. It is very important to remember that it is a direct consequence of behavior and that we want to teach a dog something, so timing is crucial.
There are four forms of operant conditioning: reinforce, this can be positive as negative and correct, this can also be positive and negative. Positive means that something is added and negative means that something is taken away. Reinforcing means that the dog will show the behavior more often while correcting the behavior will decrease.
Reinforce positive
Teaching the sit with a biscuit is a good example of positive reinforcement. If we give the dog a biscuit when he sits down, he will experience it as pleasant, the dog will see the behavior more often later (he sits more often). He will even learn that if we say “sit” and he sits down, he will get a cookie. The direct consequence of his behavior (sitting down) produces a biscuit. In this situation, something is added, namely the biscuit, hence the term positive reinforcement.
However, learning through positive reinforcement does not have to be positive for us at all, as does jumping up, for example. A sweet little puppy that jumps up against visitors will often be petted by the visitors, it is also so difficult not to. But the puppy is actually learning that jumping up brings him something pleasurable, namely attention. The jumping behavior will be shown (reinforce) more often and something will be added, namely attention. So this is taught through positive reinforcement, although it can be a negative experience before the visit if the same dog is still jumping up once an adult.
Another example is barking in front of the front door when the dog is going to be taken for a walk. We get up and take the lead, the dog becomes enthusiastic, he can go outside. Once at the front door he will bark and we know that the dog will stop barking when he is outside. So we throw open the front door as soon as possible to stop the barking and the dog has learned that if he barks at the door, he can go out.
Reinforce negative
With negative reinforcement, the dog learns that a negative stimulus for him disappears when he shows certain behavior. Teaching the seat by pushing is a good example of this. The dog will not experience pushing on the back as pleasant, but he will learn that when he sits down, the pressure is taken off. So when he feels a hand on his back he will sit more often and faster (the sitting behavior is shown more often) and the pressure is removed when he is sitting (something unpleasant is removed, so negative).
Learning that the dog should not pull on the leash with the help of a slip chain is also negative reinforcement. The dog feels pressure when he pulls and when he starts to walk next to the owner, the pressure is gone. The dog learns that walking next to the owner is much more pleasant and will do this more often, this behavior is reinforced. It is a negative reinforcement because the unpleasant feeling of the pressure is removed.
Now it is not the case that only training uses negative reinforcement. The dog is lying in the sun in the garden. The dog gets hot (unpleasant), he gets up and lies down in the shade, which is much more pleasant. Here too there is negative reinforcement because the unpleasant stimulus (the heat) disappears because the direct heat from the sun is removed.
Positive correction
It is sometimes advised to puppy owners to scare the dog by throwing something that makes a lot of noise next to him when he is biting shoes, for example. If the dog is startled, the negative experience will make it less likely to chew on shoes. Especially when he is startled every time he chews on shoes. The behavior of gnawing shoes will decrease (correction) because he has learned to be startled when he does. It has been added, namely the sound that scares him and so this is a positive correction.
Another example is when a dog sniffs at a cat and the cat lashes out. The dog will stop smelling cats in the future (it has gotten him something unpleasant) and something has been added, namely the cat’s scratch.
Negative correction
With a negative correction, certain behavior of the dog will decrease because something pleasant is taken away, the best known example of this is not coming to the command “here” or “before”. If the dog follows the command, he loses his freedom, which is taken away, and we have to make coming here very nice if he wants to keep coming. Usually the dog is put on a leash after coming and when playtime is over, we go home. The dog thus learns that when he comes, the pleasure of playing and the freedom is taken away from him and the behavior of the coming will therefore decrease.