There are three conditions which must be fulfilled here:
1. The stimulus (whatever is being done to the animal) must be unpleasant for the animal.
2. The stimulus must be linked to the behaviour you are trying to discourage - research shows that it must be applied during or within 1 second of the undesirable behaviour.
3. The stimulus must result in the behaviour decreasing or stopping.
Besides the ethical debate about whether punishment should be used to correct or train animals, the definition itself reveals why punishment often fails for purely practical reasons:
1. Many so-called punishments are not at all unpleasant for the dog. A good example would be shoving a dog away and shouting at it when it jumps up. You may think that you are punishing the behaviour by yelling and pushing, but the dog actually finds this reaction rewarding, because dogs bark at and bump each other when they play. In order for punishment to be effective, it must also be at the correct intensity to affect the dog sufficiently enough to change behaviour, but without causing unnecessary stress - this is not an easy thing to judge.
2. In my experience 99% of the time owners punish dogs long after the “inappropriate” behaviour has stopped. The dog has absolutely no idea why it is being punished at all or associates the punishment with something else altogether (usually the presence of the owner).
3. If you continually do something you regard as unpleasant to an animal without achieving the desired change in behaviour you are in fact abusing the animal.
The emotional effect of punishment
When we think about training or correcting our dogs, we tend to always think in terms of trial and error type learning or as it is known in scientific terms: “operant conditioning”. In other words we want our dogs to learn that certain behaviours have negative consequences and others have positive consequences.
However, there is another form of learning that takes place all the time on a subconscious level. This is known as “classical conditioning” or learning by association. This happens when an incident creates an emotional or reflexive response in an animal which spontaneously reoccurs when the animal is presented with something related to the initial incident.
Many people have a negative association with a certain breed of dog, because they were bitten by a dog of that breed on a previous occasion. Even though they may meet a very friendly dog of the same breed, their heart-rate will increase, their skin will prickle and they may start to perspire. This “fear” reaction is spontaneous and very often uncontrollable, because it is a conditioned response which the brain has learned subconsciously.
Dogs learn through association all the time. One of the main reasons puppy socialising is so important is that we want our puppies to have enjoyable experiences with other dogs and people (by means of play, treats and attention) so that they will form positive associations with them and therefore automatically feel good about people and dogs that they continue to meet throughout their lives. However, some of the most difficult associations to break are negative ones. A good example is the dog who became terrified of the garage, because he was trapped in the garage during a loud thunderstorm. Although the noise of the storm may have been the frightening stimulus, the garage became an object of fear by association and the dog would not enter it again.
By definition, punishment is something unpleasant. It therefore carries negative associations. However, what people don’t realise is that the associations are very seldom related to the dog’s “problem” behaviour, but to other things entirely. In this way dogs may form very strong negative associations with the owner (the deliverer of the punishment), the place where they receive the punishment (an area in the house or training venue) or some other person, dog, object or noise which happens to be present during the punishment. The more negative associations a dog has, the more emotional “baggage” he will be carrying around and the less well he will be able to learn and adapt to what is going on around him.
How Punishment can actually make some behaviour worse
Punishment attempts to suppress overt behaviour, but it does not address the root cause of the behaviour. The classic example is the dog who is aggressive towards other dogs when he is on the lead. The owner will understandably be upset by this behaviour and will want to stop it immediately. However, few people take the time to work out why the dog is behaving aggressively in the first place. The most likely reason is because the dog is afraid of other dogs, but because the lead is preventing him from moving away from the other dogs, he feels his only option to protect himself is to put on the biggest show of aggression possible, in the hopes that this will deter the other dogs from coming any closer. Having his owner beat him about the head or hang him by his collar is highly unlikely to convince him that he no longer need be scared of other dogs. In fact, he is most likely to associate the unpleasant sensations being delivered by his owner with the presence of the other dogs and be even more afraid of them in the future.
When the root of problem behaviour is fear or anxiety, punishment will only increase the dog’s stress levels and will make things far worse in the long run. Often the behaviour that people most want to punish is aggression. The problem is that aggression is an emotional response usually resulting from fear, anxiety or frustration; so again, punishment will only make matters worse. Even if the punishment is so severe that the behaviour temporarily stops, it is extremely likely that the dog will develop other serious behavioural problems.
Punishment doesn't teach dogs what to do instead
Whereas positive reinforcement of a specific behaviour increases the likelihood of the behaviour being repeated, punishment does not usually effect a particular behaviour, but ALL behaviour. Punishment causes general inhibition of responses and will make a dog less likely to try anything. This makes training extremely difficult, because a dog will be reluctant to offer new responses and so opportunities for learning become very limited.
Dogs are not people
Human beings generally operate according to a moral code. We raise our children to understand that certain things are right and others are simply wrong. If our children take something without asking permission we will call that stealing and will punish them for the behaviour accordingly. Dogs are different: they do not operate according to a moral code and do not understand what is “right” and “wrong” in human terms. Thus, if a dog takes your sandwich off the table when you leave the room it is not thinking “I know this is wrong, but I don’t care”. As far as the dog is concerned food is there to be eaten unless it is being eaten or guarded by another animal.
People often justify their punishment of their dogs by pointing out that their dogs know when they have done something wrong, because they act “guilty”. This is nonsense. The behaviour which people perceive as an indication of guilt is in fact appeasement behaviour designed to placate the angry owner. When the owner comes across some “crime” the dog has committed, his tone of voice and body language conveys to the dog that he is angry and therefore “dangerous” and so the dog tries his best to placate him through appeasement gestures like rolling over or making himself as small and nonthreatening as possible. Some owners argue that their dogs act “guilty” as soon as they arrive home, before they have even found out what the dog has done. This is a case of negative association: the dog has learned that bad things happen when the owner returns home and so acts in a nervous manner.
What do I do if my dog does something I don’t like?
There are many ways of dealing with inappropriate behaviour, depending on the situation:
1. Ignore it and reward appropriate behaviour instead - If a behaviour is of little consequence, the best thing to do is to ignore it. Training through positive reinforcement involves rewarding those behaviours which we like and ignoring all the others. Thus, if we want a dog to sit and look at us, we would only reward the dog for doing this exact thing and would ignore any other behaviour like looking away or jumping up. There is no need to punish the inappropriate behaviours because if we simply ignore them the dog will learn that there is only one way to get the reward and nothing else works.
2. Manage the environment so that the behaviour cannot occur - Many problem behaviours are very natural and normal dog behaviours and punishing a dog for acting like a dog is extremely unfair. When it comes to things like puppy chewing, pulling washing off the line or digging in the garden, it is the owner’s job to make sure that the dog simply does not get the opportunity to do these things. It is downright stupid to allow a puppy to roam the house unsupervised or to leave an adolescent dog that loves playing tug of war alone with access to the washing line. If you have a prized flower bed and a dog that enjoys digging, fence the flower bed off until the dog grows up and settles down!
3. Provide for the dog’s needs - Very often problem behaviour is simply the result of the dog trying to fulfil a need which is not being met. Hyperactivity, digging, chewing and excessive barking are very often the result of extreme boredom and social isolation. Taking a dog for a daily walk, including him in family activities, providing him with appropriate chew-toys and allowing him access to you when you are at home will go a long way towards producing a contented and calm dog.
4. Remove rewards (negative punishment) - When a dog repeatedly engages in a behaviour it means that there must be something positive that the dog is getting out of the behaviour. A good example of this is jumping up. While we may think that shouting at the dog will deter them, the fact that we are giving the dog attention is actually rewarding. Only when we remove all attention from the dog (including shouting, shoving and eye-contact) will he no longer find jumping up rewarding. Another good example is food-stealing: If every time a dog jumps up at the kitchen counter he finds something nice to eat, he will keep jumping up at the kitchen counter! If you make sure that there is never anything left lying around for him to get hold of, he will learn that jumping up at the counter is a waste of time and energy. Always try to figure out what the dog is getting out of what he is doing. If the answer is obvious, try to manage the environment so that he cannot gain the reward he is after.
5. Interrupt and Redirect behaviour - This is a good technique to use when teaching a puppy how to behave in the house. If the puppy begins to chew a table leg or other inappropriate item, distract him from what he is doing (call his name or make a funny noise) and give him something appropriate to chew instead. Again, if the puppy starts circling a spot on the carpet, quickly escort him outside and praise him for going to the loo on the grass!
6. Address underlying emotional causes - Whenever you are dealing with any problem behaviour related to fear or aggression, it is vital that you address the underlying emotional cause of the problem. If a dog is afraid of being handled, ask a behaviourist to help you to draw up a plan to desensitise the dog to this gradually. If a dog is scared of people or dogs he will need remedial socialising to overcome this.
Voice corrections - is it okay to shout at my dog?
Shouting, nagging, ranting and raving is certainly no way to train a dog or change behaviour. Again, rather try to interrupt inappropriate behaviour in a non-frightening manner and redirect the dog towards a more appropriate activity. It is also important to use “normal” tones when giving your dog cues in training. Dogs don’t need to have orders “barked” at them in order to learn and if you teach your dog in a normal voice, he will respond to a normal voice and you will not end up having to screech like a banshee when you are out in public!
There may be the odd occasion when you panic and yell at your dog out of fear for his own safety (if he is about to venture near something dangerous). If you do not regularly shout at your dog, the tone of your voice on such occasions will probably stop him in his tracks and give you a chance to get his attention very quickly. I always forgive my clients if they get a fright over something and raise their voices - the truth is that they are having an emotional reaction in that moment and there is no point in me reprimanding my clients for being human! BUT I don’t forgive clients for thinking that yelling is a constructive way of training and using this as their “go-to” method for controlling their dogs! Classes are quiet and calm, with lots of happy voices and, as a result, happy and attentive dogs!
In formal training we often use what we call “no reward markers”. These are simple phrases or words which indicate to the dog that what they have just done is not what we are looking for. Common phrases that people use in this context are “Too Bad!”, “Try again” or “Sorry, that’s not it.” These phrases are not shouted at the dog or used as a kind of punishment, they are simply a neutral sound which the dog learns to associate with the fact they will not be receiving a reward. Although the dog may be disappointed when he hears such a phrase, he will be willing to try again, because he has no fear of something nasty happening to him.
Things like coke-cans with stones in them, training disks, water pistols, dog stoppers (devices which emit a screeching sound), aerosol deterrents and citronella and shock collars are all aversive stimuli designed to stop unwanted behaviour without the dog associating the unpleasant sensation with the owner. (This is why it is called distant punishment as opposed to “personal” punishment where the owner uses their voice or body to punish the dog.) These types of aversive stimuli are very popular with many dog trainers and they allow many owners to feel that they are not directly punishing their dogs and so cannot be accused of using harsh methods.
However, the fact is that using one of these to train a dog is still using punishment and the same problems as mentioned earlier apply:
1. The owner cannot get their timing right and the dog associates the unpleasant sensation with something other than the problem behaviour.
2. The dog is not stupid and quickly realises that the owner is operating the nasty device and so the dog forms negative associations with the owner.
3. If the problem is related to fear or aggression, the unpleasant stimulus adds stress to the situation and exacerbates the problem.
4. Underlying reasons for the behaviour (fear or frustration of natural needs) are not addressed.
Very often owners become extremely reliant on one or other device and they start to use it for everything. The water pistol is a great favourite: after a while the poor dog is being squirted for everything from jumping up to failing to sit on cue. I even had one owner who squirted her dog during the recall to try to make it come to her faster - and she wondered why the dog preferred to go off by itself!
Rewards, Punishment and Formal Training
No article about punishment would be complete without a brief mention of why most trainers today have shifted away from old-fashioned force/punishment based training methods and use reward-based training methods instead.
Having originally been taught compulsive methods when I first got involved in training, I have witnessed this shift and its great benefits first-hand:
In a nutshell: Force-based training methods produced dogs that responded and performed out of fear of punishment. Dogs that were “good” at training were those that were too terrified to put a foot wrong and so acted like robots. Most dogs really did not enjoy training and only breeds that were bred to have a strong working drive were able to get anywhere at all. Many dogs learned to listen only when their owners were close enough to deliver the punishment and spent the rest of their time avoiding their owners as much as possible. Many also developed horrible associations with being on lead around other dogs.
Today, reward-based training methods have thrown the door wide open so that the vast majority of dogs can do well at and thoroughly enjoy training. Because there is no positive punishment in training, dogs perform happily and enthusiastically. They are willing to try new things, because there is no fear associated with getting it wrong. The result is a training atmosphere that is happy, relaxed and fun. This positive training environment helps to strengthen the bond between dog and owner and create positive associations with other dogs and people.
Punishment - Is it ever Okay?
With so many things having changed in the last 50 years with regards to dog training and behaviour modification techniques, I think a lot of dog owners have a vague idea that they shouldn’t really use punishment to train or correct their dogs, but are not really sure why this is. In this article I would like to discuss the definition of punishment, the problems which go along with punishment and the various ways which one can effectively and humanely deal with problem behaviour.
What exactly is punishment?
The behavioural definition of punishment is as follows: “Punishment is the application of an aversive stimulus, contingent on a behaviour, which decreases the probability of that behaviour occurring in the future.” (JSJ Odendaal, Ethology Consultancy, 2003)