We see posts from time to time (ok, far too often in reality) from the ‘balanced’ trainers, the ‘pack leader’ and ‘alpha’ theory followers. These trainers frequently say things about humans needing to be the boss in the household or their dogs are going to dominate them. They make glib comments about using ‘all the tools’ available to them and that their methods work, and work fast.
But do they? Do they really?
This is probably going to get long and technical, but bear with me. For this article we are focusing on positive punishment, covering things such as the use of ‘collar pops’, verbal reprimands (scolding), physical contacts such as smacking (or worse), and the use of equipment such as choke collars, prong collars, and electric collars applied after the dog has ‘been bad’.
 |
| Have a happy dog photo to start off on a good footing. |
Positive punishment is one of the methods available in learning theory for influencing and changing behaviour, and functions by the application of a stimulus to the dog when they are demonstrating a behaviour considered undesirable, and so reduces the likelihood of that behaviour occurring again in a similar situation. “Punishment is a behaviour change process in which a change in stimulation, during or immediately following a response, results in a decrease in the rate or relative frequency of response class members on subsequent occasions,” (O’Heare, 2017. p. 64).
This behavioural science definition of punishment as noted above differs from the general public understanding of what punishment is. The Oxford English Dictionary defines punishment as “an unpleasant experience imposed on someone for doing something criminal or wrong,”. According to this dictionary definition punishment is punitive, the result of a moral judgement. While this punitive nature can appeal to humans who can tend to take behaviour personally, as evidenced at times by dog guardians stating that the dog ‘knows what they have done’ or attribute some canine behaviours to spite, is it fair to view canine behaviour through the human moral lens?
Punishment has only truly occurred if the particular behaviour is less likely to occur when the dog finds themselves in a situation similar to the original scenario. For a stimulus to result in effective punishment, for behaviour repetition to decrease, that stimulus is aversive. An aversive stimulus produces aversion in the subject, defined by Dictionary.com as “a strong feeling of dislike, opposition, repugnance, or antipathy,”. Positive punishment means using something that the dog is known to dislike to make them less likely to choose a behaviour that has resulted in exposure to that thing previously. Deliberately exposing animals to things that they dislike and potentially find extremely distressing is, at the very least, ethically problematic. This is especially true when we consider that the use of the aversive stimulus may escalate if not effective at lower levels. “Generally, an abrupt and intensely aversive punisher, delivered immediately upon occurrence of the behaviour, consistently, each time the behaviour occurs will be more effective than a punishment procedure failing to achieve any of these criteria,” (O’Heare, 2017. p.70).
So here we have the big question on whether these methods work – does the dog continue repeating the behaviour? In the case of a ‘collar pop’ to stop the dog pulling, does it stop the pulling? A dog who receives a jab in the neck from a prong collar may walk nicely while wearing that collar, but the acid test is what happens when the prong collar comes off. Does the dog walk on a loose lead without the prong collar? If the dog pulls without the prong collar, then the behaviour has not been modified, only suppressed.
If a behaviour has only been suppressed rather than modified (stopped rather than changed) then that piece of equipment/lead jerk/smack/whatever unpleasant thing has been used to stop the behaviour being shown in that moment will remain in place/have its use continued permanently. In other words, that dog will be exposed to that aversive stimulus whenever in the situation that first prompted its use.
Some may say that the end justifies the means and that, as long as the behaviour has stopped, everything is fine. This is where a big difference between behaviour suppression and behaviour modification shows, especially when using highly aversive stimuli. Behaviour suppression stops the observable behaviour from occurring in the moment, but never addresses the cause, the ‘why’ of the behaviour. Whatever desire or emotion is underlying the behaviour will still be there and will, quite likely, become more intense as the dog has no way available to deal with that emotional state or need.
Suppressing behaviour in this way can lead to frustration for the dog and can lead to aggressive behaviour when the dog can tolerate no more. This can lead to bites occurring, perhaps to the human holding them, someone else in the same group, whether human or canine, or even an innocent bystander who happened to be passing when the dog reached that point where they could do nothing to express themselves but to lash out. Dogs have limited ways in which to communicate with humans and rely heavily on humans taking the time to learn how to read their body language correctly (which sadly doesn’t happen anywhere near as much as it should). When their communication attempts are ignored or go unheeded, what choice does the dog have but to escalate until such time as the people around them actually take notice?
The alternative to escalating behaviour and potential retaliatory aggression is upsetting to contemplate. Some dogs, when they find themselves in a situation with no escape will enter a state of emotional shutdown. Essentially, they remain still and wait for it to be over. Again, all of the negative emotions such as fear and distress are still there, but the dog no longer expresses them – they don’t work to get the dog out of the situation. The worst thing about this emotional shutdown is that many people can mistake it for the dog showing ‘good’ behaviour, that they are content, quiet, and calm when that couldn’t be further from the truth.
In recent years, a growing number of studies have examined the methods that we use when working in canine training and behaviour. Different studies have examined the effectiveness of positive punishment in comparison to other methods, others have focused on the effects training methods have on the dogs as well as the relationships between the dogs and their human handlers.
What these studies have shown is that dog training using positive punishment, whether by the use of aversive equipment such as prong collars, choke collars or electronic collars, does have massive ethical implications and that their use may cause pain and suffering (Hiby et al, 2004). This study examined reports of problematic behaviours in the dogs and the training methods used with the dogs. The study found that “the use of specific training methods may be linked to enhanced exhibition of problematic behaviours,” (Hiby et al, 2004). A correlation became clear between the number of problem behaviours reported by the dog’s caregivers and the use of methods that are based in positive punishment.
For many people who bring a dog into their family as a companion animal, the relationship they have with that dog is of importance to them. Studies have shown that the training method guardians choose to use with their dogs can have an effect on the canine-human relationship and bond (Viera de Castro et al, 2019). The study examined a number of owner-dog pairings through recording training sessions at one of six training schools. Three schools used methods including positive punishment and so came under the classification of aversive-based while the other three used neither positive punishment nor negative reinforcement, receiving the classification of reward-based. The study found that the dogs who had been trained using reward-based methods were more likely to follow their owner when the owner was about to leave the testing room than they were to follow a stranger leaving the room, unlike the dogs trained using the methods that were aversive-based. ‘This result strengthens the contact-maintenance effect shown by dogs trained with reward-based methods, with those dogs seeking contact with the owner more than with the stranger when they were about to leave the room,” (Viera de Castro et al, 2019). While acknowledging the study has limitations, the authors concluded that dogs trained using reward-based methods demonstrate secure attachment to their owners more consistently than those trained using aversive-based methods. This must be an area of concern to guardians considering their relationships with their dogs as important.
A study reviewed a number of studies which ‘examined the differences between training methods (e.g., methods based on positive reinforcement, positive punishment, escape/avoidance, et cetera) on a dog’s physiology, welfare, and behaviour towards humans and other dogs,” (Ziv, 2017. p.50). When reviewing articles based on owner-answered questionnaires, it appeared that higher amounts of fear, aggression, and problematic behaviours correlated with the use of positive punishment-based methods (Ziv, 2017). The same review article also examined studies involving direct observations of the dogs during training. One of these studies, involving military dogs and handlers, showed that “dogs that received more aversive stimuli… were more distracted and showed poorer performance compared to dogs that received less-aversive stimuli. In addition, the dogs showed a lower posture after the infliction of aversive stimuli by their handlers,” (Haverbeke et al, cited in Ziv, 2017. p.56). Another paper which focused on the welfare effects on the dog of different training methods concluded that ‘companion dogs trained with aversive-based methods experienced poorer welfare during training sessions than dogs trained with reward-based methods.’ (Vieira de Castro AC et al, 2020. P. 10.)
Another recent study looked at the efficacy of different training methods, comparing aversive techniques with reward-based training to see if there was a difference in how well the methods work. One group worked with reward-based trainers, the other with trainers deemed highly proficient with the use of e-collars. A third group was designed to act as a control group, with the trainers from the e-collar groups using methods they utilised when an e-collar was not used. The authors concluded that ‘the professional use of a reward-focused training regime… was superior to E-collar and Control Group 1 in every measure of efficacy where there was a significant difference,’ (China, Mills and Cooper, 2020 p.9). When comparing the two groups trained by aversive tool users the dogs who were trained without the use of e-collars responded just as effectively to training as the dogs who were trained with them.
There are, of course, alternatives to behaviour suppression methods – and this is where we begin to look at the ways in which we can actually change and modify behaviours.
For behaviours that are rooted in emotions like fear, we can use counter conditioning, which uses things that the dog likes to help change the emotional association they have with the source of their fear. For a dog who is scared of encountering other dogs, for instance one who has had a bad experience with unknown dogs in the past, we can pair the sight of dogs in the distance, far enough away that our dog is not reacting to them and is calm and content to observe, with a super high value tasty treat. Those of us with the sensitive and complex scared dogs prone to reactive behaviour displays know very well the value of a tube of squeezy cheese or liver paste!
In the initial stages, the dog gets the good stuff the entire time that the trigger such as another dog is in sight. At this point we are not asking the dog to do anything other than notice the presence of the other dog to get the high value food as we are looking to form a classical conditioning association. We are altering the emotional response, and this isn’t something the dog has conscious control over – hence classical conditioning, also known as respondent conditioning, which deals with involuntary responses.
What will probably happen after a time of this is that the dog will notice the presence of the other dog and will then look to us in anticipation of the good stuff being dished out. When this happens, we can begin to introduce operant conditioning into the situation and start to ask our dog to do something specific, to look at the other dog and back to us before receiving the treat. The most important thing through both stages of the counter conditioning process, both classical and operant, is that we observe the dog closely to make sure they are calm and relaxed in the situation and are not approaching threshold. I’ve previously written about signs you can watch to ensure you keep your dog under threshold, so they are not going to feel they need to react.
Often used alongside counter conditioning is systematic desensitisation. This involves starting at a level of exposure to a stimulus at which the dog is comfortable, and slowly increasing exposure while the dog remains relaxed and not showing any signs of stress. To continue the example of the dog scared by unknown dogs, once the fearful dog has begun to associate the presence of another dog with good things, we can start to close the distance between us and that other dog. It may be that progress is a single step at a time – it’s far better to aim for slow and consistent progress than to push too fast and end up potentially sensitising the scared dog further and making them even more stressed around other dogs. I also recommend moving the fearful dog closer to the other dog rather than the other dog closer to the fearful dog as the trigger moving can increase anxiety in some dogs. When we are moving the dog that we’re working with it also gives us the capacity to increase distance much quicker if we notice our dog beginning to get stressed, and so reduces the chances of damaging our progress.
What if the behaviour we want to change isn’t based in fear but in the dog wanting to do something that isn’t compatible with what we want? Pulling on the lead, jumping up at people etc.? While these behaviours are best prevented by teaching puppies from the start what we would prefer them to do instead that isn’t always possible, especially when the dog has come from a rescue centre. What we can do in these situations is teach them an alternative behaviour to the one we find problematic.
- For the dog who jumps up, teach them that they only get the attention they want when all four paws are on the floor.
- For the dog who rushes towards the front door when the doorbell rings, teach them to go a specific point where they will get something delicious when anyone is at the door.
- For the dog who wants to rush out of the front door at the start of a walk, teach them that the door doesn’t open until they can wait for you to go out quietly and calmly together.
- For the dog who pulls on the lead, reward them generously for when they are in the place you want them to be – you can use the 300 peck method – or use the environment around them to reward the behaviour so that, if they walk with a loose lead beside you, they will get to go and investigate the interesting smells over there. Alternatively, you can try stopping moving until they move back towards you and the lead loosens and then reward that by moving forward.
So, what conclusions can we draw when it comes to changing behaviour? Simply this – kind and ethical behaviour modification using reward-based methods is the most effective and ethical choice for dogs. It is the best choice for the dogs’ welfare and for strengthening and maintaining the canine human bond.
 |
| Another dog picture to clear the palette. Wish I could sleep that well! |
If you want to learn more about reward-based dog coaching, why it's the best choice, and maybe think about moving into the dog training world yourself, have a look at the courses at Canine Principles. They are wonderfully written and currently include Ofqual regulated courses from levels 2 to 5.
For more on understanding the complex fearful and/or anxious dogs, check out my book Understanding Reactive Dogs: Why Dogs React and How to Help. Link is to the UK Amazon store, but the Kindle version is available on Amazon stores worldwide (and other ebook sale sites as well). Paperback is available via a number of Amazon stores (or UK purchasers can contact me to buy direct), and the audiobook is available via Amazon, Audible, and Apple.
Sources used in creating this article:
China, L., Mills, D.S., and Cooper, J.J. (2020). Efficacy of dog training with and without remote electronic collars vs. a focus on positive reinforcement. Frontiers in Veterinary Science, 7.
Hiby, E. F., Rooney, N. J., & Bradshaw, J. W. S. (2004).
Dog training methods: Their use, effectiveness and interaction with behaviour
and welfare. Animal Welfare, 13(1), 63–69.
O’Heare. J., 2017. Science and Technology of Dog
Training, (2nd Edition). Ottawa:
BehaveTech Publishing.
Vieira de Castro, A., Barrett, J., de Sousa, L. and Olsson,
I., 2019. ‘Carrots versus sticks: The relationship between training methods and
dog-owner attachment.’ Applied Animal
Behaviour Science, 219, p.104831.
Vieira de Castro AC, Fuchs D, Morello GM, Pastur S, de Sousa
L, Olsson IAS (2020) Does training method matter? Evidence for the negative
impact of aversive-based methods on companion dog welfare. PLoS ONE 15(12):
e0225023. https://doi.org/10.1371/journal.pone.0225023
Ziv, G. (2017). The effects of using aversive training
methods in dogs—A review. Journal of Veterinary Behavior, 19,
pp.50–60.
No comments:
Post a Comment