Taking control
The ways in which we control the consequences of our puppy’s actions have changed significantly in the last twenty years or so. It is worth looking at just what those changes have involved.
Traditional dog training
In behavioural science, any consequence that makes a behaviour more likely to happen again in the future is defined as a ‘reinforcer’. Any consequence that makes a behaviour less likely to happen again in the future is defined as a ‘punisher’. An effective punisher must be something the dog finds disagreeable, an effective reinforcer must be something the dog finds highly desirable, and both must be easy to apply. Both punishers and reinforcers need to be free from harmful side effects.
Traditional dog trainers used what they thought was a fair balance between punishment and reinforcement. Unfortunately, the principles of effective training with reinforcement were not well understood, and many dogs were not offered desirable (to the dog) rewards nor rewarded in a way that creates a trained response. We will look at this more closely in a moment. But what this lack of knowledge meant was that traditional training leant quite heavily on punishment.
People’s attitudes to dogs today are very different from those of previous generations. Dogs are often considered as friends and family members, and dog owners are less inclined to punish them. Even putting to one side the ethics of using punishment, there are practical problems in its application.
Practical problems with punishment
We noted on previous pages that timing is a crucial factor in applying consequences. This can be a problem when it comes to punishment because, unsurprisingly, dogs do not like being punished. Most puppies do not hang around waiting to be smacked once they realise that you are in a bad mood. Catching and chastising a puppy where and exactly when a misdemeanour is committed is often impossible. This situation does not improve as the puppy gets older.
Regular punishment can greatly reduce a puppy’s desire to be near to his owner, and this may seriously impact on the quality of his recall. Punishment also makes the punisher feel angry when he should be remaining calm in order to think objectively how best to respond to his dog. In addition, many dogs become hardened to punishment and need increasingly severe corrections to achieve the same effect. You can see how an escalating spiral of harsh treatment can creep in to spoil a previously happy relationship, often because the owner was afraid of spoiling the dog by being too soft, and didn’t know what else to do.
Fortunately, times have changed and the way in which dogs learn through consequences is much better understood. Pioneering behaviourists and dog trainers have perfected positive-reinforcement techniques and transformed the way in which we now teach and manage our dogs. The movement away from punishment as a training tool is now widely established and training your puppy should be fun for both of you.
Modern dog training
The reliance that traditional trainers placed on punishment was not due to rewards being any less effective than aversives. It arose because trainers weren’t using reinforcement of the right kind, and in the right way.
Effective reinforcers must be highly desirable to the dog. And many dogs are not overly impressed with praise and patting. Just like you and me, one dog’s preference may be different from another’s. If you pay me in cabbages, I am unlikely to want to work for you again. Belgian chocolates, on the other hand, would be a different matter. You might be induced to provide your expertise in exchange for a white-water rafting trip, but your neighbour would sooner work for some tickets to the theatre. In the same way, some dogs will find tug games very motivating; others might prefer opportunities to retrieve a ball. Almost all dogs are impressed by certain foods – moist and highly flavoured ones are preferred. For puppies under three months old, it is also worth remembering that one of the most powerful rewards you can offer is your attention.
Studies have shown that dogs will work very hard for food rewards, and the harder the dog is prepared to work, the quicker a trained response can be created. Since food is easily portable and quickly delivered, it is an ideal dog-training tool. Using other types of reward is perfectly acceptable, provided that the rewards you have in mind are genuinely desirable to the dog in question.
The most important point is that you do not have any power to make an experience in itself reinforcing for your dog. You cannot force your dog to enjoy eating cheese, chasing balls or being cuddled. Your job is to observe and recognise what your dog perceives as reinforcing, and this requires a degree of honesty that most of us struggle with at first. We all want to believe that our dog loves praise, cuddling and being stroked by us, more than all the world. The hard truth may be that he would rather eat roast beef or chase butterflies.
Accepting that you can train a puppy far more quickly with chicken than with cuddles is only part of the battle. You also need to apply your effective rewards in an effective way. You cannot just shovel treats down a dog indefinitely and expect to maintain his new behaviour. Many people struggle with treat-based training because they don’t realise that in the long run treats must be delivered intermittently. Predictable rewards soon get boring. And bribing your puppy to be good is not the answer.
Training versus bribery
A bribe is something offered in advance in order to get the required response. If you hold out a piece of food to your puppy and ask him to sit, you are offering a bribe. Bribery is a highly ineffective long-term strategy. The owner of the dog has no control when the bribe is unavailable, or when the dog has his eye on a better reward, such as running around with your best shoes or playing with another dog.
Training is quite different. Training creates an automatic behaviour that endures, even when rewards are not always available. Dogs can be trained to respond to a signal, such as coming to your recall whistle, but also to respond to being in a certain situation, such as remaining silent in a crate, or sitting politely when visitors stroke him. Training is the process by which we create or modify a behaviour through the intelligent use of rewards. And to do this well, we need to understand how animals have evolved a way to remain persistent in the face of difficulty.
Pleasurable consequences are a great way to establish new behaviour, but they do not work as reinforcers if they continue to be predictable. This is because there is an evolutionary advantage to being persistent. Nature is not always predictable. A wild-dog family may chase five or six antelope before they make a kill. If they give up after the first chase, they would starve. Persistence in the face of intermittent rewards is essential. For this reason, dogs, and other mammals such as ourselves, are programmed to find random and unpredictable rewards more desirable than reliable and predictable ones. This phenomenon, which has been very well studied, explains our enthusiasm for slot machines and penny-pusher fairground games. It is very aptly named the ‘gambling effect’.
What this means for you is that once you have established some nice behaviours in your puppy, you will need to begin to provide rewards for them unpredictably – sometimes giving a big reward, sometimes a little one, and sometimes no reward at all. This will help you ensure that your puppy becomes addicted to good behaviour and that the behaviours you teach him become deeply ingrained.
Correcting bad behaviours
We have talked a lot about encouraging good behaviour. Positive reinforcement is a great way of training new behaviours but there are still times when puppies will behave badly. How do we stop a puppy from digging up the lawn, or jumping all over the furniture? How do we correct unwanted behaviours? Won’t we still need punishment for this?
Let’s go back to our three consequences again.
-Good outcome (reinforcement) = increased behaviour
-Bad outcome (punishment) = diminished behaviour
-Indifferent outcome (extinction) = diminished behaviour
So far we have looked at good and bad consequences, but we haven’t paid much attention to our third consequence. You may remember that an indifferent outcome, or no consequence, diminishes the preceding behaviour in just the same way as punishment. Extinction is a powerful tool and modern dog trainers are increasingly choosing it to replace the use of punishment. This means being pro-active in the way that we manage our dogs in order to ensure that undesirable behaviours are extinguished. We need to make sure that family members do not pet the dog when he puts his feet on their laps, or feed him titbits when he begs at the table. If the puppy receives no benefit from his behaviour, he will give up fairly quickly.
You can significantly hasten this process of extinguishing naughty behaviours by rewarding an alternative behaviour to replace the one you do not like. Ignoring your puppy when he jumps up at you, for example, will help him learn not to bother, but he will learn even more quickly if you also reward him for keeping all four paws on the floor.
A game of consequences
Modern dog training is a game of consequences and it isn’t difficult to play. But to stay ahead of the game you need to remember that your puppy is learning all the time, not just when you are intentionally teaching him. Consequences can work against us as well as for us. Great behaviours that you have trained can easily be extinguished. If you persistently forget to reward your puppy for coming when you whistle, he will gradually stop coming. Bad behaviours, such as yapping, can easily be accidentally reinforced by intermittently giving the puppy your attention while he is making a noise. So it is important to remember to save your attention for when he is being quiet. Good behaviours can also easily be unintentionally punished. If you immediately put your puppy on the lead after calling him to you, the recall has effectively been punished, and the puppy will soon stop coming back. Simple strategies, such as playing with your puppy for a few minutes before you clip his lead on, can make the world of difference.
Where possible, it is a great idea to reduce opportunities for your puppy to get into mischief or learn bad habits. Moving the bin into another room to prevent bin-raiding activities, and putting your precious stuff away will save much conflict.
تعليقات
إرسال تعليق