how dogs learn

How Dogs Learn

Dogs learn via the consequences of their actions. When an action results in a positive result, the dog will find it rewarding and so will be more likely to repeat it. When an action has a negative outcome, the dog will learn to repeat this action. They are always trying to achieve positive outcomes, and avoid negative ones. Because they do not have theory of mind, they are not moral creatures. They do not know right or wrong, they are very simple in this respect, there is never a hidden agenda to their behaviour.

A dog is always learning. Throughout its life, it will encounter different experiences. Their brains are flexible in that they can develop and change how they perceive things. Just because something is perceived to be a positive at one stage does not mean that that is how the dog will always see it. It could be a traumatic experience or an accumulation of mildly negative associations that change the dog’s perception. Something that the dog once enjoyed doing may become something the dog actively avoids doing because it now finds it a negative experience.

The same can be said for negative associations, so much of dog training is about building positive associations around something that the dog currently finds to be a negative.

What is conditioning?

“a behavioral process whereby a response becomes more frequent or more predictable in an environment as a result of reinforcement”

Encyclopedia Britannica

Classical Conditioning

Classical conditioning only involves automatic or reflexive responses, and not voluntary behaviour. That means that the only responses that can be elicited from a classical conditioning exercise are ones that rely on responses that are naturally made by the dog that is being trained. That the response you hope to elicit must occur below the level of conscious awareness - for example, salivation or increased heart rate.

When using classical conditioning, we pair something that is known to the dog, that reliably results in the natural response from the dog, this is known as the unconditional stimulus, with something that is unknown to the dog or neutral, that does not trigger any sort of response from the dog.

As soon as the neutral stimulus is presented with the unconditional stimulus, it becomes a conditional stimulus. If the conditional stimulus and unconditional stimulus always occur together, then the two stimuli would become associated over time. The response that was initially produced in response to the unconditional stimulus would also be produced in response to the conditional stimulus, even if it was presented alone. Pavlov called this the "conditional response."

For example, the unconditional stimulus may be chicken and the reliable response is for the dog to salivate which is its unconditional response, the neutral stimulus can be a ‘clicker’ that when introduced to the dog, triggers no response.

As soon as the clicker is presented with the chicken, it becomes a conditional stimulus. If the Clicker and the chicken always occur together, the two of them become associated over time. The response that was initially produced in response to the chicken would also be produced in response to the clicker, even if it is presented alone. This is the conditional response.

Operant Conditioning

Operant Conditioning involves performing a behaviour to obtain or avoid a consequence - intentional actions that have an effect on the surrounding environment.

Definitions to Avoid Confusion:

“Positive” = adding a stimulus/consequence NOT positive as in good

“Negative” = taking away a stimulus NOT negative as in bad

"Reinforcement” = increases the likelihood of the preceding behaviour occurring

“Punishment” = decreases the likelihood of the preceding behaviour occurring

There are Four Quadrants of Operant Learning

Positive reinforcement
Strengthens a behaviour by providing a consequence an individual finds rewarding. For example, when a dog ‘sits’ it is provided with a food reward. Increasing the likelihood of the dog repeating the ‘sit’ action.

Negative reinforcement
The removal of an unpleasant reinforcer to strengthen a behaviour. This is known as negative reinforcement because it is the removal of an adverse stimulus which is ‘rewarding’ to the dog. It requires the application of an unpleasant stimulus to the dog, in order that a trainer can then turn the unpleasantness off when the dog complies. Negative reinforcement strengthens behaviour because it stops or removes an unpleasant experience. For example, a dog that has previously learned that if it does not ‘stay’ it gets told off, when the dog does ‘stay’ it is no longer being told off. The telling off is removed if the dog completes the ‘stay’, increasing the likelihood of the dog ‘staying’.

Positive punishment

The introduction of something the dog finds negative to deter a behaviour from happening again. For example, when a dog does something that is unwanted it is reprimanded in order to deter it from wanting to do the behaviour again.

Negative punishment
The removal of something the dog finds rewarding after an unwanted behaviour is completed by the dog, in order to decrease the likelihood of the dog repeating the behaviour. For example, when a dog is asked to ‘sit’ and it does not, the reward or treat is withheld until the dog completes the ‘sit’ behaviour.