16/01/2023
A long read, but 100% worthwhile if you own a dog, train dogs, or own/train any other animals (so much applies across species).
⚠️ THE "FALL-OUT" FROM DIFFERENT DOG TRAINING METHODS ⚠️
Weighing up the risk between reinforcement and punishment based methods.
By "fall out" I refer to the negative outcomes that are a direct consequence of that training method. The things that could and do go wrong.
Sometimes my clients are worried about fall out and appear conflicted in applying reinforcement based training. They may ask me:
🗣️ “Am I reinforcing the dog for doing the wrong behaviour?"
🗣️ " Will these treats make my dog fat?"
🗣️ " Will I have to use treats for the rest of my dog's life?"
🗣️ “Am I not bribing the dog?”
Rarely do I hear concerns or am I asked questions regarding fallout associated with common uses of punishment, such telling the dog off verbally, tapping their nose or tugging on the lead.
In my opinion, humans are very quick to migrate towards punishment. Not just towards dogs, but eachother. One look online, and you see how quick people are to put one another down. We can appear to even enjoy punishing. That sense of control is highly reinforcing.
It's something I am working on, trying to ask myself: "why do I feel the need to say that?" before I actually say it. I also often edit things I have typed, pulling myself up on unconscious ego fuelled behaviour or addressing this need to punish, perhaps passively, with my choice of words. More than ever, I ask myself, "what is my motivation here?".
When punishment appears to feel good to so many, this sadly transfers over and onto the relationship between owners and dogs.
📱 Tiktok and Reels have created a platform for trainers from all backgrounds and training methods. Three years ago, from my perspective within the industry, the use of aversive methods was around but no where near as prevelant as it is now.
Today, there seems to be an abundance of trainers with zero qualifications telling people how to train their dogs using aversive tools.
Many, like myself, have felt saddened and overwhelmed by what is felt to be not just one step back in dog training, but multiple.
👊 This results in a battle of the two approaches. Reinforcement based trainers vs those who use aversives- both desperatly trying to bring the other down.
I'm guilty to having been involved, but now have largely changed my approach from trying to persuade trainers to change their methods to targeting dog owners and raising awareness.
My argument is underpinned by science, but also my own personal values and ethics. I want to be a good person!
The world is full of punishment. I fear people are habituating to it across all walks of life, including with dogs.
People are forgetting sentience and justifying dog training methods that, if used on a child or other animal, would perhaps be labelled as 'violent' or 'abusive'.
People are often focusing on making a hero out of a dog trainer, rather than noting the dog they train looks absolutely terrified or completely miserable.
Rather than focusing on what training method is most effective at addressing unwanted behaviour, because they can BOTH be effective, I want to break down which method is at HIGHER RISK of causing further undesired behaviours or a negative emotional or welfare state.
For clarity, when I describe aversive methods or use the term "aversives" I refer to:
⚡️ E-collars : these are remote controlled collars that give an electric pulse to the dog's neck, where it's worn. The intensity of the pulse (often referred to as a 'stim') can be dialled down or up.
⛓️Prong collars : these are metal collars with blunt thin metal bars that face inwards and make contact around the circumferance of the dog's neck. The collar is tightened when the lead is pulled, which pulls the prongs inwards, compressing the tissue around the dog's neck until the pressure is released.
⛓️Choke chains: the same restricting metal collar as a prong, without the prongs.
🪢 Slip Leads: The rope version of the choke chain. Can, unlike the prong or choke be used without compressing the neck due to a stopper mechanism. This is how they are meant to be used but they can be used in the same way as the choke or prong, where the collar tightens when the lead is pulled to correct the dog.
👊 “Corrections" : these are usually tugs on the lead or if you're a fan of Mr Milan, a swift kick to the ribs or jab of the hand to the throat. Sudden, aversive noises like 'tsssssk', 'pssssst' and 'ahhhh ahhh' are also corrections.
All of the aversive tools work by causing some level physical or emotional discomfort, from minor to extreme.
They stop behaviour by adminstering an pleasant stimulus that startles/scares or hurts (at varying levels) which creates behavioural change in order to avoid that stimulus. It DOES NOT work via any other means.
The level of this physical and/or emotional response is unique to the individual. Some aversives may be a mild annoyance and not really painful, while others may be emotionally distressing and/or incredibly painful.
We DO NOT get to decide and until the aversive is applied, we also CANNOT predict the experience of the recipient either. By which point, the experience has been felt and the door flung wide open for potential physical or emotional consequences to waltz in and fling their muddy boots on the table.
Subsequently, memories can now be made and learning is well underway.
Whether these memories and future expression of behaviour are in BOTH the owner and dog's mutual welfare interests will become apparent over time, rather than immediatly.
Here's an example; I'm somewhat sound sensitive, and have to often ask my partner to turn the TV down. He has no control over my sound sensitivity, especially as it varies day to day and depends on what activity I'm doing too.
It’s a minor annoyance I recover from easily (when the TV is too loud) and most importantly- the context is predictable and easily escapable for me. I know the exact source of my sensitivity and could pre-emptively avoid it should I wish (the latter is a luxury dogs do not always have when trained with aversives).
Should my partner intentionally ramp up the TV volume every time we sat down or refuse to turn it down, I’d get probably quite angry at him or try to leave. If couldn't leave I'd probably get ver stressed and would almost certainly start to feel anticipatory stress in the context (time of day, room in the house, or perhaps- even when around my partner).
Repeated, would it affect our relationship?
Thank goodness he turns the TV down without question eh?!
😬 I am always concerned by those who think they can accurately predict the full sensory experience of the recipient- even at low level intensities of aversive stimuli- and guarantee avoidance of fall out. This sense of full control brings to my mind human personality traits that are generally considered quite worrisome.
‼️You are not in full control of any living being's experience. End of.
With that in mind, when aversives are used we run the risk of the following.
1. The aversive causing physical pain, and potentially injury.
2. The aversive causing emotional distress.
3. The above being associated with the handler.
4. The above being associated with something else the dog is looking at, unrelated to the training context.
5. The above being associated with a broader context. In other words, more generalised anxiety (anticipation of pain or distress at any point).
**with 3, 4 & 5 - days, weeks or years after the punishment has been applied an emotional response can be triggered by the context or associated cues.
This emotional response MAY decrease over time, especially if reinforcers are then used for desired behaviour.
However, if the punishment maintains the absence of a previous desired behaviour it is through continued avoidance of physical and emotional discomfort**
6. The dog habituating to the punishment, resulting in an ever-increasing amount needing to be applied. This is especially the case with frequent or prolonged use of punishment, or when the intensity is gradually increased (started too low to impact on behaviour).
In other words, for punishment to work “well” there should be a singular (not multiple) perfectly timed application of an aversive stimulus of the absolute perfect intensity for that individual.
By perfect intensity I refer to it both STOPPING the unwanted behaviour and NOT causing so much physical or emotional distress that the individual is affected in the longterm.
Sounds awfully difficult to achieve to me, recalling that perception of the stimulus is not accurately predictable from one individual to another.
What started as a low "stim" may now need to be a greater electric pulse. What was a little tug may now need to be a huge yank. What started as 'ahhhh ahh' or "no" may need to become an outright bollocking.
Sounds like an awful lot of physical and emotional effort to administer from the trainer's perspective and increasingly risky business to apply.
Will the dog perhaps one day defend themselves? (This happens).
7. The final risk is "fight" behaviours, when "flight" (avoidance) of the aversive is not achievable. If we use aversive stimuli, we run the risk of being on the receiving end of defensive or redirected aggression.
It's important to note that punishment-based training with the use of aversives CAN and DOES work. But as detailed above, these risks require incredibly careful consideration should someone wish to avoid:
1. Their dog experiencing physical and emotional distress.
2. Their dog developing strong negative associations with the trigger stimulus OR other unrelated stimuli.
3. Their dog developing anxiety (stress).
4. Their dog becoming scared of them (stress and deterioration of the relationship).
5. Their dog exhibiting aggressive behaviours (stress).
🍖 FALL OUT FROM POSITIVE REINFORCEMENT BASED METHODS 🎾
1. Poor timing of reinforcement could reinforce undesired behaviour responses.
It's important to note that you cannot reinforce emotions, only behaviours. For example, giving a scared dog treats will not technically make them more scared.
2. Frustration related behaviours
Dogs may become frustrated when reinforcement stops, which can lead to undesired behaviours such as barking, jumping and mouthing. Frustration related behaviours can in some circumstances lead to more aggressive behaviours.
3. Over arousal
High intensity play games or use of high value food can elevate arousal, which can result in undesired behaviours.
4. Show me the goods
Dogs may not respond to owners without seeing what reinforcement is on offer (stuck in luring).
5. Resource Guarding
Increased high value food and items may increase resource guarding behaviours, usually in dogs with a history of the behaviour.
6. Over weight dogs
Increased use of treats may result in weight gain.
7. Reliance on reinforcers for life.
To me personally, I find these 7 risks so much less concerning than the prior list of risks. Many are so easy to work with should one have adequate skill and knowledge.
When we look at these risks and assess what might be the most effective training approach to implement, it's important to consider the following:
1. FEAR IS ADAPATIVE 😱
The brain is bias towards creating fear based memories as they will help us avoid danger and keep us alive.
2. FEAR MEMORIES ARE RESISTANT TO EXTINCTION 🧠
Fear is a VERY difficult emotion to change.
Whilst frustration can lead to threat signalling in the brain, a stimulus that elicits fear more rapid fires up circuitry in the brain than a stimulus that may elicit feelings of frustration.
3. WE ARE NOT IN CONTROL OF PERCEPTIONS
As mentioned before, we are not able to predict or manage the experience of the recipient. We can try our best through aiming to create a predictable environment and having impeccible timing. But whether we like it or not, we are NOT that recipient and we CANNOT control everything.
I consider myself a relatively skilled professional. Despite this, I do not think I am immune from making errors.
Poorly timed punishment or the wrong intensity of punishment could have life long implications. For me, the risk of making mistakes with punishment is simply too high- regardless of all my qualifictions, experience, skill and knowledge.
4. SCIENTIFIC RESEARCH 📝
Prior to research on dog behaviour and welfare, we had an abundance of research that showed the physical and behavioural effects of individuals (including rats and humans) being subjected to aversive stimuli (including varying degrees of electric shocks). As dog training and welfare evolved as a science, studies took place looking at the effects of different dog training methods.
Unsuprisingly to many and frustratingly to others, research concludes time and time again that aversive based dog training methods increase stress (physically and behaviourally) in dogs.
Research shows that this increased stress can lead to an increase in stress related behaviours, including anxiety and aggression.
Thus, research concludes- punishment based methods are a welfare concern for dogs.
With all of this in mind, my approach to behaviour is POSITIVE REINFORCEMENT BASED AND WELFARE FOCUSED.
This stance and 'statement' acknowledges that it's near impossible to go through life without doing something to a dog that will cause them some level of aversion (largely, feelings of frustration).
It is for this reason I do not label myself as a force free trainer. Pedantically, I see plenty of times I ‘force’ my dogs to do something. Although, I try my utmost best to so without causing them physical or emotional discomfort, my dogs are not free to do entirely as they please and at some point- reinforcement will cease momentarily (which can be an aversive experience if not carefully considered).
Reverting to training strategies that may be less enjoyable for the dog (such as standing still on lead, moving them gently away from something, withdrawing or preventing access to something they want) is done as a last resort- not a primary intervention.
In other words, I work very hard to minimise feelings of frustration.
Under no circumstances will I intentionally use methods that may cause physical discomfort, anxiety or fear responses to change behaviour.
When one considers that frustration is an inevitable risk of reinforcement based training- I think those who intentionally use reinforcement AND aversive tools (coined, "balanced training") should reflect on the risk of their approach causing both FRUSTRATION and FEAR.
Remembering that frustration in itself can increase threat signalling in the brain (it is a negative emotional response).
Do you really want to increase that risk of threat signalling by using aversive tools also?
Using aversive tools is avoidable.
In nearly 13 years I have not once needed to use any of the aversive tools mentioned. Large dogs, dogs exhibiting aggressive behaviour.....you name them and I've worked with them.
Busy behaviour and training professionals largely get the same dog demographic and distrubution of behaviour problems. Please do not believe the notion that some trainers get 'red zone' dogs and thus, are justified in using such methods.
When I hear aversive methods being justified because "dogs get frustrated, reinforced for wrong behaviours or fat" I think to myself.....
🤔 “THAT'S SO EASILY RESOLVED WITH AN ADEQUATE SKILL IN BEHAVIOUR MODIFICATION”
🤦♀️ I mean, if a dog puts weight on, just put it on a diet! It’s a lot easier for a dog to loose weight than it is to change complex fear based memories.
I'd rather use treats and toys for the dog's life than resort to punishing a dog across their lifespan or causing then to feel conflicted, anxious or fearful for the remainder of their life.
Pre-emptive food portion reduction, weight loss, adaptations and tweaks to the use of reinforcement (so to minimise arousal and frustration) can be a response to anticipated fall out from positive reinforcement.
Whereas, feelings of fear, anxiety, emotional or physical trauma- these things are NOT so easily resolved, even in the hands of a highly skilled professional.
Don't forget- fear is adapative and fear based memories are resistant to extinction.
How many humans have life-long counselling, struggling to overcome their fears or emotional trauma? Plenty.
Do we really expect a more primitive being, less capable (physically) of rationale, to over come their fears more easily than humans?
When you dog owners are bombarded with all of these videos of training dogs, please look beyond the described training results- because that trainer will be talking from their bias.
👀 Look at the methods and look at the dog.
Consider the potential "fall out" and how easily that could be resolved should it escalate.
Does that dog look at risk of getting fat?
Does that dog look at risk of getting a bit frustrated?
Does that dog look at risk of being physically or emotionally distressed?
Could that dog respond defensively to what the trainer is doing?
Personally, I would rather address fall out from positive reinforcement based training than I would from the use of aversives.
Not only is positive reinforcement based training more pleasant for me and the learners to engage in, it is in my opinion- lower risk all round.
NOTE: There is ZERO peer reviewed research that shows dogs trained using positive reinforcement based methods are more stressed physically (blood tests), show more stress related behaviours or more aggressive behaviours than those trained using the aversive methods detailed above. None.
This is not because all research is all bias, as is often claimed. Plenty of the research on the impact of aversive stimuli is far older than the existing debate on which training methods are best for dogs.
I I personally find that most dog owners want to have fun with their dog and want to be nice to them. I hope this post helps you all assess the risk of using gentle, kind and fun training methods on your dog and have the confidence to continue and evolve this approach.
If anyone wants to use aversives on a dog, I hope you have all very carefully calculated the cost/ risk vs POTENTIAL (never guaranteed) benefits!
PROFESSIONALS: If the dog is not our own then your duty is to so carefully calculate those risks- especially when getting owners to apply what we are doing.
An experienced professional may be able to side-step or minimise fallout when using aversives, but can they really mitigate this risk in their human learners?
Reflecting on how many owners initially struggle with the timing and overall mechanics (cue and consequence delivery) - no offense folks- I honestly feel the lowest risk is that they developed and maintained their training skills with reinforcement based methods, rather than with aversive tools.
This very last point, taking into account all the points across this very long post, is what concerns me so much about punishment based training methods and the use of aversive tools. Can owners really use it/them "well" and if not, then what could could wrong?
I personally shudder to think.
