Dogs do doggy things, they have no knowledge or concept of good and bad, right or wrong. These are human concepts. So from our point of view, dogs sometime exhibit behaviours we do what, and some we don’t want; desired and undesired behaviour.
With positive reinforcement training, or reward based training, we reward the desired behaviours – the ones we want, and we ignore the undesired ones as much as possible. We might need to remove a dog from a situation, or remove something (including ourselves) from the dog. But we do not punish the bad.
By rewarding the ‘good’ behaviours, we make it more likely that the dog will make that same choice in the future, and the more we reward it, the more we strengthen that choice.
Rewards are usually food, but can be whatever the dog in question finds rewarding; a game of tug, a stroke, eye contact. We cannot determine what is rewarding to the dog.
Sometimes we inadvertently reinforce ‘bad’ behaviours, because we forget to look at it from the point of view of the dog, and think about what they find reinforcing. For example, a dog steals a sock from the laundry, we chase them for it – dog thinks ‘great game!’ and will continue to steal socks. Or, dog is barking at the postman, we yell at the dog for barking, dog thinks we’re joining in and continues to bark.
Dogs are making choices all the time, and whenever they find a choice to be rewarding (finding a piece of food while counter surfing) they will continue to repeat the behaviour. We need to ensure that they’re getting rewarded and reinforced for all their ‘good’ behaviour, and only the ‘good’ behaviour.