By Abby Huxtable
For starters, most behaviours we class as undesirable or unwanted are totally natural to our dogs! They bark for communication, bounce when excited, have no hands and so need to use their mouths, and scavenge. However, human society perceives many of these as unwanted, so we want to change them.
We know how to prevent unwanted behaviours in the first place (see Raising a Puppy), but what do we do if our dogs form their own unwanted habits? Or come to us slightly later in their lives with unwanted behaviours already established?
This is where many may be tempted to use punishment or corrections. We do not advise either. Not only are punishment and corrections damaging to your relationship with your dog(s), but they are also incredibly hard to carry out effectively.
General consensus is that the timing of the reinforcement or punishment for a certain behaviour must be immediate for the dog to make the association with the behaviour he was performing. It’s generally accepted that you have a maximum of one second to reinforce or punish behaviour. Any longer and the dog won’t associate the consequence with the behaviour.
As we are human, we don’t always get this timing spot on. We are often a bit slow! Therefore, if using punishment to try and erase a particular behaviour, you will often not get the timing right, and your dog won’t make the connection that they are getting punished for that behaviour. This means they continue doing the behaviour and are getting punished, but they are not sure just what it is they are being punished for! Even though they are punishing the dog, the behaviour continues, so the human feels the need to escalate the punishment.
The dog is now getting punished more severely, but they don’t know what for. They just know that the human punishes them. The relationship between the human and dog begins to break down, and the behaviour actually gets worse!
However, if we use positive reinforcement, our dog is getting rewarded even though our timing may be off. Statistics and probability tell us that if you repeat something often enough, you will eventually get it right. That means that our dogs are having fun with us, getting rewarded, and eventually, we will get the timing right often enough to make the behaviour stronger.
Another method people often suggest is ignoring the unwanted behaviour. The technical term for this is ‘extinction’, and although it does work, it is very hard to do!
The idea is that if you ignore the behaviour, it isn’t being reinforced, and will therefore fade over time. However, a dog that is being ignored will usually escalate their behaviour by continuing for longer, vocalising louder, or jumping higher, for instance. This makes it much harder to ignore, so at some point we usually give in and react to them in some way, be it positive or negative. This will reinforce the escalated behaviour, as it has been rewarded! It is then even harder to ignore, as next time they will go straight to the escalation. This is frustrating for both us and our dogs.
We now know that neither punishment nor extinction are effective in getting rid of unwanted behaviours. They are extremely hard to carry out effectively, and can be damaging to our relationships with our dogs. So, what do we do instead?
The solution is to teach a mutually exclusive behaviour (MEB). This is also known as a differential reinforcement of an incompatible behaviour (DRI). This is basically an alternative behaviour that precludes the unwanted behaviour.
We can then reward the alternative behaviour, which then becomes the response to the stimulus instead of the undesired behaviour.
Note that if the unwanted behaviour is caused by fear, you would want to address the fear itself. An alternative behaviour would simply be another outlet for the fear, but your dog will still be experiencing the distress of that fear. See Guide 5 on Fearful Dogs.
Here are some examples of mutually exclusive behaviour:
- Your dog jumps up at visitors
You have been shouting ‘off’ at them every time they do it. However, they are still practising the jump in the first place and often getting reinforced by the visitor greeting them anyway:- Have them on a lead to prevent the jump. As the visitor approaches, drop a few treats on the floor, or ask your dog or a sit or down. See also: Four on the Floor – Preventing Jumping Up
- When they have four on the floor or their bottom down, the visitor then greets the dog in that position. You can also reward them if the visitor isn’t going to greet your dog
- They are getting reinforced for having all paws on the floor as they greeted the visitor, and so they learn to do that instead
- Rushing to the door when the doorbell rings
Teach them to go to their bed or place instead:- Record your doorbell sound. Play it and cue your dog to their bed or place
- Reward them when they have done so
- Repeat until they hear the doorbell and run to their bed or place
- Have a note on the door or disconnect your doorbell while you work on this to prevent them practising rushing to the door
You can find some examples of mutually exclusive behaviour in Four on the Floor – Preventing Jumping Up and Begging. When it comes down to it, you just need to decide what you’d like your dog to do instead. You can then isolate the antecedent (the thing that causes the behaviour). Then practise and reinforce the new behaviour.
It works best if you can prevent them practising the old, unwanted behaviour while you strengthen the new behaviour. In effect, you do get extinction of the old behaviour, as they get reinforced for the new behaviour instead.
Influence of delayed timing of owners’ actions on the behaviors of their dogs, Canis familiaris
Timing of reinforcement during dog training [Abstract only]
