Operant and Classical Conditioning

Operant and Classical Conditioning

Operant conditioning is a defined set of principles, developed in the 1930’s by a behavioural scientist called B.F. Skinner. The theory says we can make things happen, either good or bad, by choosing behaviours that are either rewarded or punished. We use the 4 methods Positive Reinforcement, Positive Punishment, Negative Reinforcement and Negative Punishment and where the dog acts upon the environment. Dogs will not generalise well, here, and need to be trained in all different environments in order to fully understand what is required. Once they have learned the behaviour, in the front room, we then need to train that behaviour in the garden, then in the park, and so on, until the behaviour has become generalised. E.g. the dog learns that they can make good things happen by offering the sitting behaviour. Next the dog needs to learn that those good things still happen if the sitting behaviour is performed elsewhere. We are using the Positive reinforcement quadrant here to teach the dog that sit is the behaviour we want, and we want it, to cue, in all environments.

With Classical Conditioning, we are not necessarily looking to train a behaviour, yet looking for an emotional response from our dog.  Often our dogs have learnt these responses from a single incident, and thus these become generalised very quickly. Behaviours that involve a strong emotional response where the environment acts upon the dog are classically conditioned. i.e. fear of thunder and lightning or the fear of children after an incident involving a child pulling his/her ears. Next time he/she see’s a child, they can generalise very quickly and have an emotional response of fear towards children. On the other hand, as with Pavlov’s dogs, the emotional response to the bell ringing was to salivate at the prospect of food, and dogs who have had pleasant experiences with children, soon generalise that all children are nice, and their emotional response is one of happy thoughts and not fear.

When looking to apply operant conditioning or classical conditioning when dealing with a dog’s problem, first we must define the problem. Problems are often defined as problems for the owners and not for the dogs. A dog that loves to greet us by jumping up, may have been rewarded for this by a cuddle here and there, and yet now it has become a problem, especially as the dog has grown bigger and stronger. Here we would teach the dog a new behaviour (operant conditioning) as jumping up is not necessarily an emotional response (classically conditioned) but a learnt one, probably through reward.

We would in this case use positive reinforcement, and maybe, negative punishment to change the unwanted behaviour of jumping up to the desired one of sitting when you greet me.

When the dog jumps up we can turn away (negative punishment) and when the dog puts all 4 feet on the floor, we reward that behaviour (positive reinforcement) As we continue to do this exercise, the unwanted behaviour will extinguish, and the desired behaviour increase.

Alternatively, if we do not want to use negative punishment here, we can ask the dog for an incompatible behaviour to jumping up (desensitisation of incompatible behaviours), a sit, and reward this desired behaviour. Again the unwanted behaviour will eventually extinguish and the desired behaviour increase.

If we were to encounter a dog who would bark at another dog, due to an emotional response of fear (classically conditioned), we could look to change the dogs emotional response using classical conditioning, or a technique called counter conditioning.

For dogs to learn, we must ensure that they are in the right state of mind, and a dog that is so fearful may not be able to learn, as its emotions are such that it is not cognitive to learning. This happens if we subject the dog to too much of the stimuli that causes the fear in the first place. Some trainers will flood the dog here in an attempt to prove to the dog that there is nothing to fear and subsequently change the dogs emotional response and learn to be happier in the presence of other dogs. This method, in my opinion, can also cause a dog to shut down and not be able to learn that other dogs are nothing to be fearful of, or worse still lead to some aggression towards the other dogs. If we ensure that the dog is in a clam state before we introduce the stimuli, then this helps towards learning. We need to change the dogs emotional response from fear by adding something nice in the presence of the stimuli. We can use high quality treats to feed the dog, ensuring that our dog is not overly stressed at each stage. We will keep the stimuli(other dog) at a distance, where our dog is happily eating the treats, thus now pairing the presence of another dog with something nice. We can move the stimuli closer over a period of time or over different sessions, as our dog becomes more at ease, subsequently classically conditioning the dog to enjoy the presence of other dogs.

 

Jeff Sasse (2010)