What is the difference?
As trainers who train using mainly positive reinforcers to reward behaviour that we like or want, it is important to identify what things our dogs find reinforcing. It is the dog that decides what is reinforcing and not the handler. Rewards are defined by the dog not the trainer.
Positive reinforcers can be broken down into two categories; intrinsic reinforcers and extrinsic reinforcers.
Extrinsic rewards are external to the dog. They come to the dog as rewards for displaying a behaviour from an outside source which is usually the handler; food and praise and petting would be extrinsic rewards. We use food rewards all the time when teaching the dog a new behaviour. For example, every time the dog sits in response the “Sit” cue he gets a treat. In terms of learning theory the frequency of the sitting behaviour should increase.
Intrinsic rewards on the other hand, come from within. The dog derives satisfaction and reinforcement by being given the opportunity to display the intrinsically reinforcing behaviour. For example, for a retriever the opportunity to retrieve an object would be intrinsically rewarding.
How would we use intrinsic rewards in our training?
Initially when teaching a new skill from scratch, it is often more convenient and more efficient to use a marker such as the clicker, and a high rate of reinforcement. Once the behaviour is learnt and is under stimulus control, then only occasional rewards are required to maintain it. Very often, depending on the skill taught and the internal motor patterns of the dog which are determined by the breed, the skill or behaviour itself becomes rewarding.
When training my retrievers in the field and provided the learned behaviours have been trained and proofed using clicker and food, my dogs know that the reward for good behaviour is the opportunity to display the intrinsically reinforcing behaviour of the retrieve.
Most retrievers have an inbuilt desire to chase, retrieve, or just carry an object and our challenge as retriever trainers is to design our training programmes to use these natural patterns of behaviour as rewards.
For myself, I find that I will work on a solution to a problem because the challenge of finding a solution provides a sense of pleasure. This sense of pleasure and achievement is enough to motivate me to tackle the next problem with a sense of expectation, irrespective of any external reward that may be forthcoming. Of course, the external rewards of trophies and awards is also very reinforcing, but this is not what drives me to solve the problem or to become a better trainer of dogs. My pleasure comes from the task itself.