Some dog behaviourists/trainers believe dogs should be trained using only positive reinforcement. Here is why I don’t agree with this – and why I believe that by adopting this ‘positive only training’ approach, we are in fact killing dogs with our kindness.
Before going into why I believe positive reinforcement only training, allow me to briefly look at operant condition, the heading under which most behavioural work (as well as some obedience training) ultimately falls. Without boring you with too many details, operant conditioning essentially has four different, very strong consequences.:
- Something GOOD may be presented or start to increase a behaviour. This – POSITIVE REINFORCEMENT – is related to pleasure, like, for example, treats, games or lovely long runs in the park, woods or a field.
- Something GOOD may be taken away or end to decrease a behaviour. This – NEGATIVE PUNISHMENT – is about taking something the dog likes (i.e. a treat or favourite toy) away.
- Something BAD may be presented or start and a behaviour subsequently decreases. This – POSITIVE PUNISHMENT – is about ‘handing out’ direct punishment like, for instance, using check/choke chains or electric collars or hitting a dog.
- Something BAD may be taken away or end to increase a behaviour. This – NEGATIVE REINFORCEMENT – is about removing something that can cause distress and pain, like stopping to hit a dog or press a shock collar’s button.
What Makes a Trainer “Good” or “Bad”?
In my opinion:
- Most good trainers and behaviourists would, except under certain circumstances, tend to only use a positive reinforcement/negative punishment combination.
- Most bad trainers/behaviourists would regularly use all four or just one of the above methods/consequences.
You may find my saying that positive reinforcement could be bad very strange. However, many trainers and even a few so-called behaviour experts believe that the exclusive, isolated use of positive reinforcement is THE one and only ethical way of changing dogs’ behaviour.
If these people fully understood what training and behavioural methods are about, they would think again before taking the work of B.F. Skinner (creator of the behavioural quadrant; learn more) and removing 75 per cent of it., According to Skinner, the term ‘positive’ simply means the presentation of something shortly after a response was made – and that presentation then results in an increase of the intensity or frequency of the response.
The ‘positive’ in what we call ‘positive reinforcement’ neither has anything to do with what’s good or bad, beneficial or detrimental, nor is it in any way connected to what anybody wants/does not want. Positive reinforcement could, in fact, by its very definition include what could be conceived as harsh punishment.
If, for example, a dog snarled and threatened you and you responded by kicking and abusing the dog, that abuse would most likely result in the dog intensifying his aggression (increasing the level and frequency of his aggression) towards you. Your abuse will therefore have functioned as POSITIVE reinforcement. Similarly, treating/praising a dog at the wrong time can positively reinforce bad behaviour just as much as punishing him the wrong way can.
Understanding behaviours and operant conditioning are more complex by far than the somewhat simplistic belief that reinforcement is good, and punishment is bad. Oh, how I wish things were that easy – you know, a bit like in the old Westerns where heroes always wore white hats and the ‘baddies’ invariably wore black ones…
Positive Only Trainers
To be honest, I would defy any behaviourist or trainer to demonstrate their ‘positive reinforcement only’ methods. Training any animal/mammal – be it a dog or a human; a killer whale, dolphin or horse – using only positive reinforcement is totally impossible. These individuals are either lying or deluded, because let’s face it, simply:
- Closing your fingers around that treat your dog is about to snatch is NEGATVIE
- Turning your back on your dog when he jumps up is NEGATIVE
- Walking your dog on a leash is NEGATIVE PUNISHMENT
Many of these self-proclaimed positive only trainers actually teach in colleges/universities, marking the exam papers of students supposed to become the behaviourists and trainers of the future. Not only do these people teach, but they frequently mark students’ papers based purely on their own, unsupported and biased theories and opinions – and I have experienced this first hand…
Over the years, I have taken many behavioural courses. On one occasion, the senior tutor at a renowned Animal Training College (an individual in charge of writing and setting exams at this college!) not only completely ignored every book we were instructed to read, but actually disagreed with their content – and subsequently marked our papers based on her personal opinions.
Excuse me for saying this, but I found this totally idiotic and utterly useless: students are given books to read & study and their expertise and knowledge of said reading material is then supposed to be tested by them having to answer several related questions – yet their tutor ignores the books and marks their papers for their personal latest ideas and beliefs. It’s no wonder people are giving up on education…
Thankfully, on this occasion, many students had their marks reversed after making complaints. One can, however, not help but wonder just how many good students fail their exams because of their tutors’ madcap ideas!?
Quoting Skinner, these people talk about operant conditioning while at the same time completely ignoring the plainest of facts – namely that operant conditioning is not restricted to a single, isolated element. In my humble opinion, these people – who believe they are so fair and kind – are probably responsible for far more dogs being sent to rescue centres or euthanised than any of those really bad trainers whose training methods involve lots of aggression and positive punishment.
In my experiences, positive only trainers fail to understand that dogs, like all animals (and humans!) must be aware of causes and consequences. In their haste to show how kind and caring they are, they completely ignore the main principles and tenets laid out within ethical conditioning. We cannot claim to use friendly, kind methods while using negative punishments under operant conditioning – which is precisely what would be described as the achievement of a lead. What’s more, we cannot pick and choose what suits us from Skinner’s work and claim to have the “moral high ground” when clearly, this is not the case.
Think about it: your dog is enjoying himself running around and then you attach a lead. What happens? You instantly have NEGATIVE PUNISHMENT – because you have taken something your dog enjoys (his freedom) away. Do you imagine for a moment that your dog considers this a positive experience? I often wonder how many dog owners have noticed that their dogs walk to heel like a dream when they are off the lead – yet as soon as that lead is on, they pull like runaway steam trains. Has anyone ever wondered why that is?
Many prospective members were failed by the Association of Pet Dog Trainers (APDT) simply for saying “No” to badly behaved dogs. The APDT believes this stresses dogs. So, where, I ask myself, is the fundamental basis for learning and discipline? How is it possible to instil an ability to understand unacceptable actions’ consequences and mutual respect? You are being rebuked just for saying “No”!
The APDT was founded by brilliant behaviourist John Fisher, who sadly passed away in 1997, just two years after founding the association. I knew and highly respected this remarkable man, who was instrumental in helping me understand that coupling training expertise with the behavioural path was the right route to follow. It pains me to think how he would feel if he could see what these raving lunatics have done to his organisation.
Should Bad Behaviour be Ignored?
These are the so-called training experts who will tell owners to just turn their backs and fold their arms when their dogs are biting them. A response that ultimately could cause the poor dogs yet more anxious and frustrated – and subsequently more aggressive and dangerous. When initial attention seeking behaviours and frustrations fail to work, the result could all too easily be much more substantial escalation of unacceptable behaviours and/or aggression.
I am often asked to step in after such positive only training classes. Going back to scratch, I then put a kind, yet effective program that will actually benefit both dog and owner into place. Astonishingly, some of these trainers and behaviourists “treat” dogs without ever even touching them.
Instead, they will sit and observe the dogs for hours before writing 57-page reports on what the owner should do – while being utterly incapable of doing the work themselves. It’s like asking someone to learn to drive from a book. I have calculated that more than 40 per cent of my work consists of working with dogs other trainers/behaviourists failed to make a noticeable difference with. Almost all these cases involve trainers keen to point out that their methods are force free or positive only.
Instead of teaching dogs to heel, these people put devices they consider “kind and gentle” – like harnesses or, worse still Haltis – onto the poor animals. Have you ever seen these poor dogs’ desperate struggle to get those hellish contraptions off their faces? I have – it broke my heart to see their hair rubbed right down to their skin; to see the blisters, abrasions and infections caused by these infernal devices riding up and rubbing against their eyes… They say these devices are a kind alternative…
All Haltis, head collars and harnesses work on restraining dogs to stop them pulling. Surely, they can only achieve this through discomfort and pain. Doing nothing but restrain dogs, they also do not teach them to heel. If your “training aids” fail to actually train a dog – i.e. if your dog starts pulling again the moment you take the darn thing off – then it definitely isn’t a training aid. It is something that must work using discomfort and pain.
These torturous devices were not created by Gandalf or Harry Potter – there’s no magic here. So, how do you imagine they could possibly work? Looking at the operant conditioning model, these wicked devices would definitely come under the heading of positive punishment – because they cause distress, discomfort and pain. The method I teach actually trains dogs to heel, rather than simply restraining and distressing them.
Many recent media and press articles called for changes in the way families, schools and organisations instil discipline and ethics in our children. Many experts have stepped forward and said that by not teaching and commending the many virtues of respect; by not instilling a moral and ethical compass in our children, we are failing recent generations.
Unfortunately, many positions within education and training are held by people who believe that children and pets should only ever be controlled by positives – nothing else. No negatives. No Discipline. No Control. Reward the good and ignore the bad.
We are sadly creating a society failing to instil decency, morals and self-control. When even saying “NO!” is not acceptable in a society, anarchy will reign. We owe both our children and pets educational and training programmes that help them understand right from wrong, guidelines and boundaries, respect instead of contempt. Cruelty, distress and pain are NOT required.
Killing Dogs with Our Kindness
Far more pets are being euthanised for training and behavioural problems now than ever before. Why do you think this may be happening? According to a report I just read, the main reason for our dogs being euthanised is deviant behaviour. I would firmly place this statistic onto the doorstep of all those so-called force free and positive only trainers.
The Importance of Consistency
All animals, and especially mammals, need tracks to run on – and consistency is without doubt the key. Every action has a consequence in the wild, and youngsters are taught these boundaries by pack members – mothers, siblings, alphas and controllers. Surprisingly, in packs of coyotes, jackals, wild dogs and wolves, very few injuries are caused by aggression and fights.
The powerful, strong sets of rules & hierarchies of these animals filter right down even to the pack’s bottom members. Their ethos of belonging is exceptionally strong and pack positions and ranks are both respected and strictly adhered to. This is because the packs are not dictatorships run by alphas, they are family units – father, mother and offspring.
Understanding that in these situations, leadership is rarely tyrannical but is instead based almost entirely on mutual respect is very important. Ranks are normally reinforced by posture aggression, rather than real attacks. There is an obvious reason for this: overtly aggressive leaders and tyrants would not engender respect, but fear; they would not instil confidence, but insecurity. For their hunting forays to be successful the pack must work as a cohesive, efficient unit.
As humans, we cannot be our dogs’ alphas (we are, after all, not dogs). We can, however, lay out guidelines, boundaries and rules for them to follow – and we can also control their resources (i.e. food, games, toys, even their access to us). Being their resource controllers enables us to gain mutual, fear-free respect and subsequently create deeper bonds with our pets. Just like in any other community, family or pack, there must, however, be guidelines – and these guidelines must be clearly defined to work.
The Role of Stress
The role of stress in learning is vital. Excessive stress means learning collapses, while inadequate stress means no learning occurs. Shortly after our decision to launch the Profession Association of Applied Canine Trainers (PAACT) – an organisation relating to combined dog training and behaviour – I was contacted by a lady belonging to another organisation. She would not disclose which organisation she was a member of, but I have a feeling it was probably the APDT.
This lady demanded to know whether our members were permitted to stress the dogs we treated. Immediately realising that I was dealing with a rank amateur without any idea about how humans and other animals learn, I explained how important stress is in learning and that stress can be good or bad. I pointed out the importance of certain aspects of stress in both the mental and physical growth of dogs.
I also gave her an example: for puppies, learning and stress does not begin the day they arrive at their new home, but they day of their birth, even as early as while they are still in their mother’s womb. Born blind, deaf and incapable of smelling anything, pups have just two senses until they are approximately 2 weeks old: taste and touch. At around 10 days, their sense of smell begins to work; their hearing kicks in at around three to four weeks and they can finally see properly at about six weeks (they can see a little before that, but it’s a little as though they are peering through a veil).
Humans carefully and gently handling a puppy during this all-important time (from their first day of life onwards) can generate mild stress responses within the pup, which helps the pup to grow both in physical and emotional terms. This is therefore good stress.
Puppies that were handled during their lives’ first few weeks grow and mature quicker, are more resistant to diseases and infections, and are more stable in general. Handling everyday-stress much better, they are also more curious and exploratory and learn faster by far than puppies that were not handled during this vital period.
This is why it is so very important to be careful when choosing a puppy. It is best not to buy pups from large breeders or puppy farmers, who simply do not have enough time to handle pups – and you should never get a pup from pet shops. It is best for breeders who clearly love dogs, have them indoors and, not being commercially minded, are not just ‘in it’ to make money.
Not surprisingly, the lady never got back to me with an answer or acknowledgment. To be frank, I didn’t expect her to. All people who do not really understand how pups/dogs learn do is to blindly follow the latest theories and fads. This, unfortunately, is nearly always to the dogs’ detriment.
Let’s Stop Killing Dogs with Our Kindness
We must realise and understand that dogs – both puppies and adults – must be given a full spectrum of experiences. This means giving them not only the positives, but also consistency and a clear understanding of which behaviours are acceptable, and which are not. Clear indication of what is right or wrong makes it far easier for dogs to understand and follow direction – and makes them much calmer and more settled. Resulting in a much deeper, stronger owner/dog relationship, this ultimately benefits both dog and owner.
Let’s hope that the number of pets having to be sent to rescues, re-homed or even euthanised will decline once again because we start actually training them, rather than just showering them with praise.