While it is human nature to discipline or punish somebody for what they have done wrong, we should understand that pets are not people. They don’t comprehend things on the level we do, and they don’t understand the difference between right and wrong. Just as there are many ideas on how to ideally discipline children, there are various opinions on how to train dogs to do what we want them to do.
Conflicting information available
If you do even a small amount of research on dog training methods, you will discover a lot of conflicting information. Start speaking to different people, and you will find similarly contradictory advice. One book will suggest time-outs for unwanted behaviour, while another will recommend using only affection as reward; a website will condone shock collars for dogs who aggress, while the next will propose psychotropic medications; your friend will tell you about their trainer who only uses treats, and your work colleague will say treats are ineffective.
It’s extremely difficult to know where to turn for reliable advice – information is freely available, and everyone has an opinion on how dogs should be trained. To add to the maelstrom of confusion, dog behaviour modification and training is an unregulated industry, meaning that any person can advertise their services regardless of their knowledge, skills, qualifications or experience – there are no standards that have to be met, and no code of ethics to abide by. This is a dangerous situation for both dogs and their people who simply need help. The only escape from this mess is to find the most reliable information possible.
There is a lot of controversy in the dog training world regarding the use of punishment. Some advocate the sole use of punishment, some prefer a combination of punishment and reward, and others will only use reward. Controversy is unnecessary though, because in the case of punishment, there is reliable information available.
Scientific studies on punishment
Multiple scientific studies have been conducted on the use of punishment and negative reinforcement in dog training, and the results are emphatically clear – reward-based training and the sole use of positive reinforcement are the way to go.
Punishment itself is a disputed term, and many people disagree on what constitutes a punisher. Technically, what most of us understand as punishment is actually referred to as positive punishment. The ‘positive’ aspect of it means that something is added rather than removed, and the ‘punishment’ aspect means that whatever is added causes a behaviour to decrease in frequency, making it less likely to occur. In sum, a bad consequence causes a behaviour to diminish (it may be called a ‘correction’ or ‘startle’, but the definition confirms that these are punishments).
This can be contrasted with negative punishment, in which something desirable is withheld in order to diminish a behaviour.
Reinforcement has two facets as well, both negative and positive. Negative reinforcement is a tricky one: a behaviour is strengthened by the avoidance of something unpleasant, while in positive reinforcement, a behaviour is strengthened by the addition of something rewarding.
To return to punishment, it seems like a highly effective means of stopping a behaviour. You just add something unpleasant and the behaviour stops. But it’s not that simple, unfortunately. In order for a punishment to be effective – to stop a behaviour occurring again – it has to meet certain criteria:
- The dog must find the punisher unpleasant, and this is a very personal trait. Some dogs enjoy the spray from a water bottle; some might think a knee in the chest when they jump up is part of a game. If they find the stimulus enjoyable, it is going to positively reinforce the behaviour instead.
- The punishment must occur immediately after the behaviour, and every time the behaviour occurs. If it’s not immediate, the dog may associate the punishment with another behaviour. If it doesn’t occur every time, the dog may risk the behaviour having learned that punishment is not always forthcoming.
- The intensity of the punisher has to be high enough to stop the behaviour, otherwise the dog will become resistant to it, but low enough not to cause negative fallout.
- The punisher must be associated with the behaviour, not with the trainer or the environment. The dog may learn to fear or mistrust the person, environment or indeed any other aspect of the situation with which he associates the punishment.
- The punishment must suppress the behaviour. If the behaviour continues, positive punishment has not This is an integral part of the definition of punishment. If punishment has not occurred, then something else has – ineffectively applying negative stimuli to a dog constitutes abuse.
So, if you want to use punishment effectively, you must have high-speed reflexes and timing to punish at the precise moment, you must know the exact intensity of the punishment required to stop the behaviour, you must implement the punishment every time the behaviour occurs, and you must ensure that the dog associates the punisher only with the behaviour that you want to stop. Sound difficult? It is. Achieving all of these criteria is extremely challenging verging on impossible, and the case is similar for negative reinforcement. There is a lot of risk involved and ample scientific evidence has confirmed the dangers of using punishment in dog training.
A 2017 review of scientific papers on dog training methods concluded that the use of aversive and punitive methods negatively affected the welfare and behaviour of the dogs involved in the discussed studies, and that these methods were no more effective than reward-based methods. In fact, studies have found that the use of punishment may worsen behaviour, and is especially associated with increases in fear, stress and aggression.
One study, for example, found that 11% of dogs displayed aggression in response to the use of a prong or choke collar, 15% aggressed when the handler shouted “No!” and 43% aggressed when hit or kicked.
Punishment and negative reinforcement are distinct risk factors when it comes to dogs displaying aggression, including barking, lunging, growling or biting, towards members of their own family – this is extremely concerning, as many families with dogs also have children. The use of punishment and negative reinforcement affects your relationship with your dog: a study comparing positive reinforcement and negative reinforcement training methods on two groups of dogs displayed this very graphically. The dogs in the negative reinforcement group hardly looked at their owners, displayed lip licking and yawning (behaviours associated with stress) and also avoided their owners more frequently than those in the positive reinforcement group. This not only implies an effect on the human-dog bond, but will necessarily also influence the results of training, as stress inhibits learning.
This is by no means an exhaustive review of the available scientific literature, and entire books can (and have) been written on the topic. For example, there is so much research on the hazards of using shock, vibration or citronella collars that it is impossible to include it all.
The bottom line is that punishment and negative reinforcement have the effects of causing discomfort or nervousness at best, and fear and aggression at worst. They are exceedingly difficult methods to implement effectively without fallout, and don’t confer any additional benefits to training. Dogs are likely to learn to distrust people, and become avoidant or defensive. Aversive methods alter a dog’s behaviour, but not always in a way that is desirable or in the interest of the dog’s welfare.
We owe it to our dogs – and our cats, horses and other animals – to ensure that any approach we take with them is based on sound, evidence-based information. We have the tools to train our dogs and modify behaviour problems without resorting to methods that are unpleasant because of this information. There is no need to use even the mildest “No” or a stern gaze when teaching our dogs. And the result is a dog who wants to learn, who wants to behave appropriately, and most importantly, who has complete trust in his people.
This is all a bit confusing, so let’s simplify with an example of a common behavioural complaint. A dog jumps up on people for attention when they arrive, and to address this, one of the following learning quadrants can be applied:
- Positive punishment: The dog is smacked on the nose when he jumps up. Next time a visitor arrives, he does not jump up. The smack is something that has been added to decrease the behaviour of jumping.
- Negative punishment: The dog is ignored until his feet are on the ground. The dog wants attention and it is not given until he stops jumping. In this case, something the dog wants is being kept from him in order to decrease the jumping behaviour.
- Negative reinforcement: When the dog jumps up, the person holds his paws (something the dog does not like) until he stops jumping. The dog is trying to avoid something unpleasant – having four feet on the ground is more likely to happen as a result. This is similar to positive punishment in some ways, and the two can easily be confused.
- Positive reinforcement: The dog receives a treat when his four feet are on the ground. Something has been added (the treat) to make a behaviour more likely to occur (standing).
 Ziv, G. (2017). The effects of using aversive training methods in dogs – a review. Journal of Veterinary Behavior, 19:50-60.
 Herron, M. E., Shofer, F. S., & Reisner, I. R. (2009). Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. Applied Animal Behaviour Science, 117(1), 47-54.
 Casey, Rachel A.; Loftus, Bethany; Bolster, Christine; Richards, Gemma J.; & Blackwell, Emily J. (2013). Human directed aggression in domestic dogs (Canis familiaris): occurrence in different contexts and risk factors. Applied Animal Behaviour Science, 152, 52-63.
 Deldalle, S., & Gaunet, F. (2014). Effects of 2 training methods on stress-related behaviors of the dog (Canis familiaris) and on the dog-owner relationship. Journal of Veterinary Behavior: Clinical Applications and Research, 9 (2), 58-65.
By Katherine Brown