Lets talk about how everything we do, affects how our dogs react to us and the world around them.There are two ways in which dogs learn about the world around them: through classical (Pavlovian) conditioning and Operant conditioning.
Classical conditioning is a dogs natural learnt response to a trigger. For example:
- You open a treat packet (makes rustling sound),
- They turn to look at you,
- You give a treat.
Initially your dog doesn’t relate the rustling sound to anything, but they quickly pick up on that, but soon your dog hears the rustling of a packet and they look at you, which you then reward. How many times have you opened a packet of crisps and your dog is right in front of you looking at you and waiting for that treat? I bet it happens all the time. Even after you remove the reward the trigger sound (rustling of the packet) still makes your dog display that behaviour (watching you).
Operant conditioning is when you knowingly add or remove something to make a dog do a behaviour more frequently, or stop doing the behaviour completely.
Positive Reinforcement & Negative Reinforcement both make a behaviour more likely to happen.
Positive reinforcement introduces a reward to make the behaviour more likely to happen: For example: You give the dog a treat every time they sit, so you make the behaviour (sit) more likely to happen.
Negative reinforcement takes away an aversive (something bad) away to make a behaviour more likely to happen. For example, when you remove the pressure of the choke chain when the dog sits, so you make the behaviour (sit) more likely to happen because the dog reacts quicker in order to remove the discomfort of the choke.
The same can also be said for positive and negative punishment.
Positive punishment introduces an aversive to make a behaviour less likely to happen. For example, you use a shock collar to add an electric shock (an aversive) to make barking less likely to happen.
Negative Punishment takes away a reward to make a behaviour less likely to happen: for example: You stop giving your dog attention (something good) to stop him from barking at the door. In this case, the attention he considers good, is that you open the door for him each time because he is barking, by not opening the door, you are removing the reward.
There are many more examples in our daily lives where both operant conditioning and classical conditioning are in play. This is how animals learn. You can teach almost any animal using these two principals, including us humans.
Now that we know the basics of how us animals and our dogs learn. We can better understand why our dogs do the things they do. Simply put, the behaviour they do has a desired consequence, or not. Positive reinforcement and negative punishment go hand in had quite well. We usually unknowingly do this on a daily basis. This is why so many dogs can have behaviour problems – because we are teaching them to act that way without even knowing that we are doing it.
For example: if my dog jumps up at me all the time, and I stroke her when she jumps up at me (because I love the affection she shows) I am rewarding this behaviour, because at the time she jump on me, the reward (or consequence) is that she gets lots of cuddles. If I was to remove the reward (cuddles and attention) and simply ignore her when she jumps on me, she will start to look for other ways to get my attention. Initially it might seem like she is getting the hang of it because she calms down a bit. But then suddenly there will be a burst where she might jump on me more and more and maybe even nip or whine or bark to try to get my attention. If I continue to ignore her, she will then realise that doing all those extra behaviours don’t work, and she will eventually give up, possibly go to lie down. I can then go to stroke her and cuddle her while she is lying down. I am now rewarding her for being calm and relaxed instead of the unwanted behaviour of jumping up.
Dogs often exhibit a sudden change in behaviour called an “extinction burst”. When we try to change a behaviour there may be occasions where the dog will suddenly go back to the old unwanted behaviours and maybe add some extras along the way, as a way to “test” us. If we give in and reward them (by getting frustrated and giving in to their demands to be stroked), then they have learnt that they can behave in such a way including the extras, so now your dog is doing the newly reinforced behaviours which are jumping and nipping and barking. But if we are persistent, and ignore the bad behaviour burst, then the chances of your dog having another extinction burst is reduced more and more until they no longer try to “test” us. Eventually the behaviour we want to phase out will no longer be a problem because the dog has learnt that it gets them nothing.
This is of course not going to work when there are environmental factors that act as an automatic reward to the dog. If you are teaching your dog not to jump on people, but when you take your dog to the park, they jump on everyone they pass, they are automatically rewarded by the interaction they get from the passer by (even if the passer by is disgusted by your dog, your dog still finds it amusing and therefore its still a reward). In this case, applying the same principle as above is going to be pointless because the chances are that your dog is not going to get bored from lacking attention in this situation. In these cases, you have to combine management, positive reinforcement, and negative punishment to help teach your dog what you want it to learn.
Positive punishment is when a you add something bad to make the dog less likely to do something. Sometimes you might shout at your dog to stop them from running into the road. The dog finds shouting unpleasant and so will stop in its tracks startled by your loud voice. Another example could be if your dog has a bone in its possession, and you make a loud sound with a broom to make them release the bone. By association, the dog will relate the broom, and you as a bad thing.
Its important to understand that when we do use positive punishment, we are adding more stress to the dog, which can weaken the bond you have with your pet and the relationship they have with you and your environment. In sensitive dogs you may actually cause them to withdraw themselves from you and they will want to avoid the situations that cause them stress. In dogs that are tougher to train, by increasing the punishments out of frustration to force the dog to do as they are told, you may actually end up creating an aggressive dog.
In the scenario I describe above, if I use positive punishment to stop my dog from jumping, I would smack my dog each time (+ aversive) to stop her from repeating the jumping behaviour. Because my dog is very sensitive, I would find that she would now have learnt to avoid greeting me at the door completely and hide in a corner being scared. This is not the sort of relationship I want to have with my dog.
When training your dog, its more effective and creates a longer lasting bond if you use positive reinforcement. The more you use methods that don’t stress your dog out, the less you will have to resort to more forceful methods to solve problems.