Reinforcement vs punishment: A dog trainer's perspective.

Introduction 

The role that punishment versus reinforcement should play in dog management and training is turning into a major area of friction and debate, being played out globally, in research and in popular media. The discussion has spiralled into a heated separation of the two camps – being the “positive reinforcement methods only” versus the “dominance and punishment (using aversives) are essential” cohorts. It might be more accurate to use the distinction of "force-free vs coercion-based training", however, folks do misunderstand the word "coercion", in that they think it means simply convincing a dog to do what you want - using bribery, petting or leash pressure - rather than the strict definition of "persuading someone to do something by using force or threats". Either way - two camps have separated and these groups are diametrically opposed, often characterising the other camp as tribal, manipulative, myopic and simply ignoring the scientific and natural evidence – as well as ethical and moral arguments about how to behave[1] and telling the other side what they should/should not do.  It doesn’t seem to get any easier, let-alone resolve and is quite confusing for the novice dog guardian hoping to do the right thing in teaching their dog good behaviours and even preparing for the dog sports. The aim of this article is not to perpetuate the myths, but to attempt to analyse the areas of difference, define some points of agreement and, perhaps, point to a resolution pathway for the two warring camps.

In short-form, just to set some foundations, modern domestic dog training relies heavily on the studies of Pavlov[2] (1839-1946: Classical Conditioning), Thorndike[3] (1874-1949: Law of Effect in Learning Theory) and Skinner[4] (1904-1990: Operant Conditioning). This work led to a dependency on the psychological model of “Behaviourism” (as separated from other important psychological models such as Cognitivism, Constructivism, Socialism and Experientialism)[5]. These works were deepened and brought forward by Skinner’s co-workers, the Brelands[6].

Domestication and training of dogs can be argued to have started with hunter-gatherers as much as 10,000-20,000 years ago, but in more recent times we can see formal task and result training using positive food reward training referenced in 1760[7] and positive reinforcement as obtaining better results than punishment and fear to be traced from as early as 1848[8] in sporting and gun dog training. However, after the turn of the 20th Century, a base of military dog training (WWI/II – Most and Koehler)[9] and then Hollywood/Disney movie training for animal performers, played an important role. The key methodologies successfully used in these latter scenarios were forceful, compulsive-dominance training techniques. The “do this when commanded or be punished” did work – dogs did perform the tasks required. This methodology was seemingly nicely coupled with the “alpha male wolf” hypotheses, established through studies in the Basle Zoo in the 1940’s (Schenkel[10]). That hypothesis was later reaffirmed in 1970, but then overturned by the year 2000 (Meck[11]) as based on a false and misleading premise of captive wolves from disparate groups being thrust together in a confined area – totally unnatural compared to the free and wild wolves which operate as family groups over a broad domain. Unfortunately, not only did this whole “dominance pack theory” and the apparent attraction of labelling male dominance behaviour as “alpha” get out of hand, it is still touted by some dog trainers as the way to train dogs. “You have to be the alpha pack leader to get your dog to obey you” is a common statement based on a false and thoroughly debunked premise.

Over the last 20 years, to maybe as far back as the last 50 years, a realisation has dawned that animals trained in this forceful-dominance-punishment manner (referred to as compulsive training, punitive/coercive training, avoidance learning or escape learning[12]) are not only being harshly (in many cases, cruelly) dealt with, but they are not developing their potential as willing companions and they are not encouraged to use their natural talents to think for themselves and develop a keenness for learning[13]. Surely these natural talents and cognitive characteristics are what is really desirable in a companion animal – and indeed in a military or police partner.

Notwithstanding the modern enlightenment, there is a recent trend to train a dog using “leash pressure techniques” – a pretence that there is something magical about how a savvy trainer uses the leash. This simply comes across as terminology obfuscation, as the methodology is actually reverting to the compulsive training techniques, but pretending it is really something else, which only a sophisticated dog trainer can comprehend. Despite the options and such new notions – all of which, admittedly, have some value in understanding the debate, if not in advancing the training outcomes - the pendulum has swung towards a greater awareness of methods described as “positive reinforcement training” (referred to as “positive-only”, “force-free”, “fear-free”, “inducive training” and elsewhere as “motivational training”[14]). In parallel, variations, step-wise/half-way measures and clarification of training models have emerged. Although a little theoretical in application and seemingly (at least at first reading) allowing for the use of aversive training methods[15], these can be viewed as sensible attempts to reconcile the opposing sides for the advance of animal handling. Examples are LIMA: Least Intrusive, Minimally Aversive[16] and LIFE: Least Inhibitive, Functionally Effective[17] which is critical of LIMA as a "handbook of aversives" rather than positive reinforcement techniques - LIFE focuses more on animal welfare (The Five Domains) .

As much as the old ways of dog training will not go away completely, the new ways (as with many societal shifts in understanding psychology, learning development and even human social interactions) have swung to some extreme points. This is not terrible – it seems a common phenomenon that in an effort to counteract an established and well-rehearsed view, an alternate view – possibly a diametrically opposed view - needs to pass through a stage of being “over stated” – the extreme polar opposite - just to draw the clear distinction and trigger a rethink. Hopefully, a true balance and continuously improving management and teaching methodology will emerge – but it hasn’t yet, even though most of the evidence for advancing a better way appears to be available[18].

One caveat on what follows. The discussion and views expressed are aimed at companionship dog management and training (commonly referred to as “pet dog training”) as well as training for the dog sports of Obedience, Agility, Hoops, Frisbee, Dock Diving, Gun Dog Retrieving, etc. Does it apply to the very special military and police dog training of high drive dogs, such as the Belgian Malinois? Well, yes – and no. Yes, in the general terms of reference as to the ethics and benefits of positive reinforcement training and teaching dogs to think and respond[19]. No – in terms of the details of what these special purpose dogs are required to learn to overcome/combat and the actions they need to perform in their special jobs. This is part of the evolution of the whole dog training regime and will be briefly revisited in the Discussion and Conclusions.

Clarity in confusion

Some points of confusion in the current debate are rather frustrating, as pundits from both sides stumble over them, trying hard to prove their own methodologies of teaching dogs as right or supreme – the only sensible way. This is more than simply semantics, although certain distinctions in definitions will be described as important. So, following is a list of key issues, misunderstandings and debates in the dog training world today, coupled with reflections on their veracity or applicability.

A. Important Definitions

1. Sweeping statements about dog training methodologies and tactics are only as good as they really do apply to all dogs – all breeds, all ages and all individual personalities[20]. That is rarely the case. Yes, there are fundamentals about puppy upbringing and obedience trial training[21] and so on, but it is important to have regard to dog breed and development stage of the dog (puppy, adolescent, adult, geriatric) as well as the individual animal’s previous life experiences. All modern breeds were cultivated for a reason – companionship, hunting, coursing, scent work, guarding, protection, etc., which means that dogs have various levels of drives and senses – and hence a range of responses to stimuli. Then we complicate these fundamentals with cross breed mixes, often messing up behaviour characteristics and predictability. We like to think that one size fits all – but that is only true for some elements, some of the time. Setting out to train a Malinois, Newfoundland, Short Haired Pointer, Rottweiler, Border Collie, Asian Shepherd, Scottish Terrier and Cavalier King Charles Spaniel or Italian Greyhound all exactly the same way in every element is doomed to failure on so many fronts. Breed characteristics and individuality need to be recognised.

2. The next important definition to establish in discussing dog training methodologies is this: are we discussing “obedience/manners/dog sports/trick training/establishing a behaviour pattern”[22] or is it “change of unwanted behaviour”[23]? This might seem obvious and is sometimes acknowledged, but is just as often overlooked - causing arguments to be at cross-purposes - and it is key, because the focus is different and the management, persistence and even “intensity” of training methods one might choose to apply are different – especially if in the behaviour change scenario there is a safety issue to consider. Now, there is definitely a blurred overlap between obedience/ manners/trick training and behaviour change, but to keep it simple, a straight forward guideline might be: 

For obedience/manners/ trick training, positive reinforcement techniques are not only all that are required, but achieve the best outcomes for the trainer, the dog and the relationship/bond/trust between the two.

In behaviour change, the argument becomes more convoluted, because a highly reactive dog may require some constraint and redirection – which the positive-only trainers generally regard as “aversive” and therefore not acceptable. Positive-only trainers go so far as to state that it is preferable to euthanise a dog than to use aversive methods to re-program its behaviour.  This seems not just a silly cop-out by a trainer who throws up their hands and gives up, but obscene. The logic appears to assume that all aversive methods are akin to torture (the extreme interpretation) hence killing a dog (albeit humanely) is preferable. Hopefully readers can see the skip in logic here – killing can generally be regarded as rather aversive[24]. This is not providing an excuse to use aversives, it is simply that the extreme argument does not help the understanding, let alone common dog welfare.

3. The next definition involves that fuzzy term “aversive” – and no, it’s not as simple as “aversive means aversive”[25]. AVSAB (American Veterinary Behaviourist Society)[26] define Aversive in the dog behaviour context as “anything unpleasant (emotionally or physically) that is used to decrease an unwanted behaviour”. Examples of an aversive may include verbal reprimands, pushing an animal into a submissive position (alpha rolls, dominance downs) threatening body language, shaker cans, spray bottles, citronella collars, leash corrections, choke chains, prong collars, or shock collars. Interesting - and sort of covers the field (albeit highlighting the extreme end) – but is this really useful for people trying to come to grips with training and managing their dog’s manners/obedience or behavioural issues?  Does this position inadvertently support euthanasia of a misbehaving dog because the only other option is torture? A terrible and (surely?) unintended consequence.

There is, furthermore, a timing element in defining an aversive action. If a dog is pulling on the leash and the guardian slows or stops and gives a cue such as “Steady” or “Wait” to alert the dog that the forward motion is slowing down or stopping, in that moment of time is an aversive being applied? Well, yes, certainly. The methodology of training a dog to desist leash pulling is discussed in another article on this site (“Common Dog Training Challenges”) but for now just recognise that there are auditory cues, delivery mechanisms, timing and duration options – as well as severity – in aversive actions. No – this is not acknowledging that a quick jerk on a prong collar or buzz on an e-collar is therefore OK because it is brief – hopefully readers can see the difference (and follow-up by referring to the Training Challenges article). 

B. So, what do dog trainers mean by aversive training techniques - and are they consistent in these definitions?

1. Some positive-only/force-free advocates will not allow a trainer to use a “non-reward marker”, such as saying “no” or even “ah, ah”, “oops” or “wrong” to a dog. Now – remember the point above about the nature of what is being discussed – if training a new obedience element or trick, this “never say no” discipline can readily be adhered to – the trainer simply keeps revising and repeating the instruction/ demonstration/luring (possibly shaping and/or using "chaining" to build the outcome) until the dog responds, then goes with the happy voice and reward (the clicker or voice bridge and praise, treat or game). Force-free proponents find it rather hard to defend an argument that telling a dog “Oops” in a quiet voice when it performs the target stunt incorrectly causes pain, anguish or lasting stress in the dog. Whether a non-reward marker is really necessary and helpful to the training outcome is another question. If the trainer has established a strong bond of trust with the dog, the dog will be every bit as delighted as the trainer with the final successful outcome and everyone is in a positive and happy place, without even the mildest aversive.

2. However, if a safety issue is at stake and the guardian desires a dog to “leave it” (that rotting rubbish in the park) or “no – don’t growl at that child”, or “don’t lunge towards the road side” (with passing cars the obvious danger) or the dog starts to react unfavourably to a passing dog or person, then it would be a very unusual person who does not utilize a “mild aversive” to quickly curtail the undesirable activity. That might be saying “No” or “Ah, ah”, "Leave it" (if already trained), leash control (not jerking a prong collar/zapping a shock collar and screaming) or “spatial pressure” (such as “body blocking” the dog – standing between them and the trigger or object of reactivity - not kneeing, poking, punching, kicking or hissing at the dog, as certain popular TV dog practitioners like to do). That is, containing the forward movement of the dog and diverting attention/adding space/lowering the intensity to diffuse the situation. Are these strictly classified as aversives by the religiously positive-only advocates? Yes, they are. But, remember, that these are all effectively safety situations – as well as re-programming opportunities if positive reinforcement for the desired outcome is added to the equation. That doesn’t mean the trainer has failed as a positive reinforcement trainer, has reneged on their training ethics and “gone over to the dark side”. The suggested “reprogramming training” for each of these challenges are summarised in a Training article on this site ("Common Dog Training Challenges").

3. Part of the confusion in the whole “aversives” discussion is that positive-only advocates generally seem to assume that aversive means extreme leash jerks, yelling, poking, kneeing, kicking, alpha rolls, choke chain collars, sound blasts, slip leads, prong collars, electric shock collars, no-pull harnesses and even nose harnesses (i.e., head halters – which, surprisingly, even some positive-only trainers are OK with – yuck!) – all elements designed to cause suppression, discomfort, pain, stress, fear, or all of these in progression or at once. Unfortunately, the debate is fuelled by the evident and frequent use of such tools by TV/YouTube-popular dog trainers today, who, amazingly, earn a profitable living abusing some very sad dogs in need of real help.

C. Arguments commonly made in favour of aversive techniques

What do those who are absolutely ardent about the need for aversives argue? Generally, the arguments in favour of aversives run along five basic lines:

1. Stress and fear are a natural part of life and are therefore not just OK, but a fundamental need, a “life requirement”. The arguments go like this:

i. “Rule of the jungle” – reaction to and ability to handle/comprehend aversives are bred into the animals DNA; understanding aversives (stress, danger, pain) are a fundamental of survival. However, consider that the dogs in the training debate are domesticated animals, not wild and free-roaming, living in a “kill or be killed” environment. Humans have cultivated dogs as pets, companions and working partners and have thus taken on the role of provider and protector - Rule of the Jungle simply does not apply.

ii. Aversives are essential - “we humans learn not to touch a hot pot on the stove after just one burning incident” – therefore (the logic seems to go) dogs need to find out not to do certain things in the same way, viz., by experiencing a painful consequence. The ridiculous nature of this argument should be obvious – adult humans can distinguish between hot pots and cold pots, but dogs cannot, so applying this type of correction (literally) would simply make a dog fear all pots[27].

iii. Drawing parallels between human logic/emotions and those of dogs – especially prevalent in describing the way humans react to sports or games in which some stress or pain might be or is inflicted (martial arts, sky-diving, competitive physical sports, schoolboy’s hand slap game[28], etc). This method of argument may seem logical on the surface, but is seriously flawed, because of the differences between humans and animals. Humans (well, some humans, at least) find the aversive elements and adreneline flux in certain stressful activities exhilarating, fun and developmental and, therefore, the argument goes, there is no harm causing stress or pain to a dog through aversives in training – they will find it exhilarating and learn “lessons of life”. But these parallels between human behaviour are generally not at all illustrative – and commonly just plain misleading. So, as an example – a fair comparison to the stresses (and for some, competitive fun) in the hand slap (also called “hot hands”) game, would be that the trainer says to the dog “OK – here’s the learning and developmental game – it’s called “hand slap” - I try to hit you with my hand – if I succeed, I get to hit you again, as hard as I like. If I miss, you get to bite my hand – if you succeed, you get to do it again - as hard as you like”. Ridiculous. People understand the rules of the games they play and have a choice – they can enter the game or sport or not, call “time” when they have had enough - and they can bask in the psychological or physical development they feel they have gained from the experience of testing themselves against another human or the natural elements. It doesn’t work that way for dogs – they do not have the human cerebral dynamic - we anthropomorphically so often assign to them - to comprehend such a game and learn from it, or be exhilarated by it (not to say that all forms of anthropomorphic parallel are taboo[29]).

iv. Stress is a good thing, so aversives used in training causing stress are not a problem: the argument, apparently, is that when going to the gym, an athlete puts stress on their muscles, heart and mind and become stronger – and live a longer healthier life – because of stress. Without trying to regurgitate 20th Century psychology texts[30], this simplistic argument just misaligns the types of stress being discussed and waylays the discussion. Not helpful. Exercising a dog, doing agility training, scent work, tracking, water recovery, gun dog training, etc., all put the good physical and mental stress on the dog akin to the human in the gym (or playing chess) – this is not the issue of stress induced through fear, pain and uncertainty.

2. The dog trainer wants a quick result and doesn’t have time to fluff around with these wishy-washy, nice-guy methods of positive-only training. We are told that “in real life, people don’t have the time to spend on these positive reinforcement techniques to change unwanted behaviours or walk somewhere the dog’s behaviour won’t be triggered”. This argument is troublesome because it seems like in this push-button, fast-paced age, we need instant outcomes, as if we really desire an “App on our phone” to hit so the dog responds as required {Hey – maybe an electric collar delivering a compulsive command via electric shock would do that!? We can definitely market that idea to the gullible public looking for the quick fix in all of life's challenges and make a fortune!}. Dog training simply does not work like that – the dogs are not robots. Guardians need to appreciate what is involved in owning/caring for a dog – it takes time, effort, persistence and patience. No one should think its easy.

3. The subject dog is trying to dominate the guardian and fight to be the alpha leader of the pack, so aggressive dominance-compulsion is required as a counter-action to boss this out of the dog’s psyche.  Now – if the trainer runs around on all fours and urinates on trees the subject dog is sniffing at the time – maybe there is something in this? Not. The topic of “dominance” is the subject of a further article on this site ("Canine Dominance - is it a Real Thing?") – for the purposes here the issue is that the quick-fix, aggressively aversive, remedies compulsive trainers couple with this piece of bad logic do not lead to long term behavioural cures or solid training outcomes. The most these harsh methods achieve is behaviour suppression or inhibition in that instant. Nor, very importantly, do they establish harmony and a bond of trust between dog and trainer/guardian[31]. Compulsive trainers believe that a dog must be forced to confront its fear, dominance or aggression and learn that the human is boss – the pack leader. This argument is so often seen and even verbalised, the naïve could be led to believe it is a real thing, when in fact these remedies are simply applied as a form of “flooding” (or bullying/provoking) to generate and then suppress behaviour, without allowing for any mental or instinctual reaction reset opportunity in the dog. Learned helplessness is real[32] – just witness it happening on hundreds of YouTube videos marketed by trainers who employ this form of compulsive punishment training to “change behaviour” – invariably heralding the process as curing the problem in a single session.

4. The “four quadrants of operant conditioning”[33] should be applied by a balanced dog trainer with equal regard. The catch in this argument is that Skinner’s and Thorndike’s conclusions regarding reinforcing behaviours through consequences, although often quoted, are rarely understood or explained. There are two points to consider:​

i. The 4-quadrants[34] are not really equally balanced (as has been taken up by labelling “balanced dog trainers”). This confusion presumably arises because the Skinner “teaching options” are commonly represented as four equal quadrants of a circle or equal quarters of a square graph. They were never intended to be utilised in some kind of equal proportion (as the label “balanced” infers). Rather, they illustrate an academic model to categorise certain processes in an effort to clarify what might be considered in the “Behaviourist” model of Skinner’s and Thorndike’s groundwork.  Remember the adage that “all models are wrong; some are useful”[35].

ii. Skinner and Thorndike really concluded that positive reinforcement (that single quadrant/quarter in the model) is the most effective, lasting teaching methodology, which actually encourages further learning. 

5. If a dog in training is only told/shown what to do and never what not to do, then the training will be somewhere between confusing, unfair, ineffective and even aversive – hence stressful. This argument is also seriously flawed.

On the surface, this appears to be an appealing argument, according to the way humans think. But does a dog think like that? In the absence of having Dr. Dolittle conduct a survey, the answer to go with is “probably not”. Given a chosen “cue” (body language or auditory) to request an action or behaviour, there are many things a dog could do instead of what is targeted (including doing nothing) – but if they are not rewarded for any of those actions/inaction, there will be no reinforced desire to repeat them. Adding into the training some demonstrative elements of body language, copying, redirection (such as pointing the desired target), luring/encouraging, will all offer opportunities to teach the dog that this particular targeted action is the way to be rewarded – either by a treat, game or just happy and exuberant praise from the trainer. What the dog can then deduce is that repeating the correct/desired behaviour outcome will achieve the target result - a reward – no other action (or inaction) by the dog will deliver that.

Discussion and conclusions

The often referenced "science proves that", or "based on the consensus of scientific studies/opinion"[36] and even "evidence based training", normally in the context of attempting to weaponize science or chosen evidence to demonstrate either that positive-only or use of aversives in dog management and training is absolutely right, are misleading and often unsubstantiated statements. These quotes misunderstand what science is and how it works - and how a good scientist should operate. There is a real danger in claiming “science-based training methods” – if only because both sides of the debate make the same claim – equally vociferously. Both are on shaky ground when attempting to prove absolutes in behavioural psychology.

A scientific concept, model or theory is an attempt at an explanation of certain observations under certain specified (constructed or natural) conditions. The empirical observations (the hard data) need to be separated from the inferences and deductions. Data need to be repeatable and observable in varied environments/situations, understanding how observations might be skewed by external contingencies (or even observer bias) – possibly not even noticed or appreciated at the time of data collection, but only reflected upon by subsequent research and thoughtful analysis of alternatives[37].

Then, it must be appreciated that the resulting findings are what can logically (scientifically) be deduced at that time, with what is known and must be subject to future modification with additional or varied data or observations, more knowledge reflected from parallel branches of science or psychology, or greater insights into the possible influencing elements. This does not make the original deduction completely wrong and useless – indeed, in many cases, had it not been for that earlier outcome, deduction or opinion, the next advance in the science might not have been prompted to occur. Think of each scientific deduction, made with the best intentions of understanding and furthering the knowledge pool, as akin to rungs on an infinite ladder – each rung plays its part in providing access to the next rung. Think of it in terms of "It's not that the last researcher was wrong, its just that I think I have a better interpretation or method to resolve this challenge - so I'm glad that researcher triggered the idea". We might think we are there and know it all, but we are not at the top of the ladder.

Further, attempting to link unrelated scientific evidence/conclusions from disparate fields, experiments, scientists, animals, times, etc., to prove an unrelated over-arching hypothesis is fraught with dangers and highly likely to mislead. There must be a commonality – a ribbon to tie the data and results across the differing fields, studies and outcomes. This is not only challenging, but very commonly simply doesn’t exist and is probably a linkage and conclusion the original scientists never intended.

Confirmation bias and cherry picking – basically the light the commentator is shining on the subject - do influence what is interpreted from the “science” and what becomes popularly believed, according to the social mores trending and the leanings – and level of understanding - of the author/reader/practitioner. The challenge is to keep an open mind and see the options and variations in applying any model – the case in point here being the success and impacts of dog training methodology. Application of a methodology is a good way to test a hypothesis, bearing in mind the subject of the trial application and the observed response or outcome, as well as the ethics and good animal husbandry of any trial. Most importantly, the long-term impact of a dog training method needs to be holistically and thoroughly understood – in the case of aversives anywhere on the scale, it rarely is.

On the subject of applications to test a dog training methodology hypothesis – this is already done hundreds of times on TV and YouTube and the outcomes of aversive methods have been covered here. A spurious and naïve argument sometimes put forward by “balanced trainers” (especially those who clearly favour dominance training) to finally solve this question of what’s best, “positive-only or balanced”, is a challenge that two trainers utilising opposing training techniques should run a head-to-head trial to see whose dog sits, downs, recalls, or modifies an unwanted behaviour fastest, to prove their technique is the best. Really? No problem. Just find two dogs identical in every way (breed, age, personality, physical and mental state, identical level of training and with identical behavioural issues, if that is the target) in the same place in their heads on the day of the trial, with the same familiarity with the environment the test is conducted in, then after the trial check the dog’s ongoing welfare and the willingness and longevity of the dog’s responses to cues. Sorry – failure on all counts. At the very best it would be a "splitting hairs" competition and, at worst, would be a waste of time, because the test would need to be repeated with different identical (or at least extremely similar) sets of dogs and trainers multiple times (hundreds or thousands?) to validate the results – and the evidence is already visible. The suspicion is that this challenge is put forward in the knowledge that it will not be taken up – because it is a childish, unanswerable and silly concept.

The influence of the closed mind[38] (and confirmation bias – only liking what one already thinks) invades these positive-only versus balanced and/or dominance-aversives debates at a high rate. The hope is that a better methodology – a reckoning – will come through listening to the other side and appreciating the huge variation in dog natures, behaviour examples, management and training strategies - and outcomes. And no, we are not looking for winners and losers here. The approach that will stymie getting to where all dog owners and trainers really want to be, is that ideas, evidence and opinions are totally disregarded without analysis or understanding because of “user bias” (preference or prejudice based on incomplete information) and, as a consequence, not even considering an alternative view, nor attempting to understand exactly how another dog trainer really works. For example, what does that trainer mean by balanced? Does that mean they use force and punishment and prong collars on all dogs, all the time, as a first choice, or something different? Do they, in fact, have any useful ideas or methods which a positive-only trainer might consider? How does a positive-only trainer act? Does that mean they constantly shower a dog with treats until it gets sick? Do they, in fact, have any useful ideas or methods which a balanced trainer might consider? Most likely neither of those extreme characterisations of the two sides of the debate is valid, but they are very often the priority assumptions. Admittedly, the extremes we witness in training – particularly with aversive proponents – does make the mind shrink and be closed to certain trainers, but that should not be allowed to impact the entire debate. Don’t let the lunatic fringe stop sensible discourse, new ideas and learning.

On the introductory caveat regarding the views expressed applying to military and police dog management and training, the carve out for these specially trained working dogs does not mean that the use of certain techniques or tools (otherwise regarded as aversive) in these disciplines is therefore universally acceptable in any circumstances, which the dominance/compulsive trainers might leap on with an “Ah ha! So, aversive techniques and punishment are allowable or even required!”. No - it means that there are special requirements in these fields of dog training which the average dog trainer just does not get exposed to. So – an open mind is required. There may be more to this than what we can readily see, because the public are not privy to the details and the published research in this field is limited, especially real-life boots-on-the-ground studies (for obvious reasons) [39]. There is, however, a notion that a modern-day military/police dog trained with positive reinforcement techniques is more likely to “think for itself” when confronted with an unrehearsed situation[40], a field rich in further research opportunities[41].

To end where we began, a realisation has dawned that animals trained in the previously popular forceful-dominance-punishment way are not only being unethically (in many cases, cruelly) dealt with, but they are not developing their potential as willing companions - and they are not encouraged to use their natural talents to think for themselves. Moreover, such compulsive training methods overlook the flow-on effects – on the dog and the dog-guardian relationship[42]. Surely this relationship, bond and trust are what we really want in a companion animal – and indeed in a military or police partner? Further, are aversive training methods really required when they are applied, or do they just satisfy the frustrated handler looking for a quick result?

The saving grace in all of this is that dogs can prove to be incredibly resilient and affectionate companions, despite human misdemeanours against them, especially if the groundwork of providing the fundamentals for survival (shelter, food, water) plus bonding and trust are in place[43].

In conclusion, the pendulum has swung towards a greater awareness of methods described as “positive reinforcement” – but - the pendulum is still in motion. The definitions used in describing training and behavioural change are important, just as is distinguishing the target outcomes of dog manners/ obedience or reprogramming behaviours, all in the context of the dog’s breed, personality, age and life experiences.

Could any of the views expressed herein change with new information, ideas and insights? Absolutely. The cynics will be thinking “Huh, this author doesn’t have the courage of his convictions – backing both horses – can’t make up his mind!”. If anyone is thinking that, then I feel sorry for you. Though the leaning towards positive reinforcement and avoiding aversives in all scenarios is clearly very strong, new insights, logic and the open mind rule.

References

[1] Ethics and morals are sensitive subjects, and people grapple with accepting what others state as obvious, because ethics are codes and behaviours set by societies (groups) and morals are the property of the individual (based on a personal compass). So, whilst a set of guidelines might be proposed (and seem self-evident and right to one group) neither ethics nor morals – whilst possibly generally agreed across some cohorts - can be universally set down and there is no compulsion to accept them by all players in the debate.

[2] Pavlov, I. P. (1897/1902). The work of the digestive glands. London: Griffin: (1928). Lectures on conditioned reflexes. (Translated by W.H. Gantt) London: Allen and Unwin.

[3] Thorndike, E. L. (1911). Animal intelligence. New York: Macmillan.

[4] Skinner, B. F. (1938). The behaviour of organisms: an experimental analysis; New York, Appleton-Century: (1951) How to Train Animals; Scientific American, 185(6), 26–29.  (1976) About Behaviourism; Vintage books (Random House) NY; 291pp. Note that although Skinner is the most often quoted authority of "Behaviourism" in animals, the the founder of behaviourism psychological theory is actually regarded as J. B. Watson (1879 - 1958) - see review by Roegiger, H. L. III (2004) at URL https://www.psychologicalscience.org/observer/what-happened-to-behaviorism

[5] Not intended as an exhaustive list. Experienced and observant dog trainers will often wonder what role these other models might play and when they become relevant, even overshadowing Behaviorism in modelling a dog’s reactions to learning. A complex subject and opportunity for further research.

[6] Breland, K. and Breland, M. (Introduction by Bailey, M.) (1996) Animal Behavior. Republ (2018); Storymakers Inc.; 352pp.

[7] Fairfax, T. (1760) The Complete Sportsman: or Country Gentleman’s Recreation; Reprint paperback (2021) Legare Street Press; 248pp.

[8] Hutchinson, W. N. (1848) Dog Breaking: The Most Expeditious, Certain and Easy Method, Whether Great Excellence or Only Mediocrity be Required, with Odds and Ends for Those Who Love the Dog and the Gun; 4th ed; London, John Murray. Republ. (2013) Vintage Dog Books; 430pp.

[9] Most, K (Colonel) (reprint 2001) Training Dogs: A Manual; original 1910, translated to English from the German 1954; Dogwise Publ.; 214pp. Koehler, W. R. (1962) The Koehler Method of Dog Training: Howell; republ. 1996 Hall & Co. USA: 378pp (Koehler trained a number of dogs for Disney movie roles).

[10] Schenkel, R (1946/1948) Expression Studies of Wolves: Captivity Observations; Basle Zoo and Zoological Institute of the University of Basle; Translated from German; 58pp: (1947) Behaviour; 1 pp. 81-129.

[11] Mech, L. D. (1970) The Wolf: The Ecology and Behaviour of an Endangered Species. Natural History Press (Doubleday Publishing Co., N.Y.) 389 pp. (Reprinted in paperback by University of Minnesota Press, May 1981): (1999) Alpha status, dominance, and division of labour in wolf packs. Canadian Journal of Zoology 77: 1196-1203: (2000) Leadership in Wolf, Canis lupus, packs. Canadian Field Naturalist; 114(2):259-263:

[12] McLeod, S (2024) Operant Conditioning: What It Is, How It Works, and Examples; URL https://www.simplypsychology.org/operant-conditioning.html#. Gadbois S (2015) 51 Shades of Grey: Misuse, Misunderstanding and Misinformation of the Concepts of “Dominance” and “Punishment”; Do you believe in dog? April 2015; URL https://www.doyoubelieveindog.com/2015/04/

[13] Burch, M R (2022) The Evolution of Modern-Day Dog Training: Today’s dog trainers owe much to their predecessors; Dog Savvy Los Angeles; URL http://www.dogsavvylosangeles.com/blog/2022/8/13: provides an excellent historical evolution summary and commentary.

[14] Acknowledging that motivation can be induced in a being by either positive or threatening/punishment means – the intended interpretation in the nomenclature here is clearly positive reinforcement motivation. A further, somewhat derogatory, term, which misses the point, is “treat training”.

[15] A highly experienced and level headed trainer is needed to actually apply these somewhat complex methods as proposed by the authors (quoted below) because each methodology requires sophisticated and cautious decisions as to when to undertake the next step (especially in the LIMA protocol). The concern is that inexperienced practitioners (or those with a leaning towards the aversive methods) will become impatient and ramp up to the next level of aversion or inhibition on the basis of the action being justified as the next “least aversive or most functionally effective” action, rather than persisting with the lower steps, just because the model seems to say “it is allowable”. Note, however, an important caveat and a key understanding of the LIFE Model is that "functionally effective" actually means ethical training techniques which have worked through behaviour change by relating to/focusing on the cause of that (undesirable) behaviour - i.e., not just effective by any means (such as severe aversive application to stop/inhibit behaviour).

[16] LIMA: Lindsay, S. R. (2005) Handbook of Applied Dog Behavior and Training: V. 3, Procedures and Protocols; Blackwell Publ.; 795 pp.; pp. 29 & 726; See summary and explanation at CCPDT website; URL https://www.ccpdt.org/about-us/least-intrusive-minimally-aversive-lima-effective-behavior-intervention-policy/; Also see Article "Should certain dog training devices be banned" (this website) with reference to Lindsay's (op. cit.) position on electric collar use on dogs; and the Training document "Training equipment used (and misused) in dog training" (this website) - this latter emphasises that "equipment" does not train a dog - the human trainer does that - a mistake so many, unfortunately, make. Also Friedman, S.G., (2008). What’s wrong with this picture? Effectiveness is not enough. Good Bird Magazine, 4:4, 12-18. (2022) Why Animals Need Trainers Who Adhere to the Least Intrusive Principle: Improving Animal Welfare and Honing Trainers’ Skills; PDF at BehaviorWorks.org; URL https://www.behaviorworks.org › files › articles; a criticism is that the concept justifies (or promotes) aversive tools (the criticism is not without reason but, perhaps, not entirely fair, as this not seen as the key objective).

[17] LIFE: Fernandez, EJ (2024) The least inhibitive, functionally effective (LIFE) model: A new framework for ethical animal training practices; Journal of Veterinary Behavior; 71; pp.63-68; This model differs from LIMA in several ways as it attempts to clarify/overcome what are seen as failings in LIMA . The focus of LIFE is increasing desirable canine behaviour through what a trainer (and the canine) should do. For the Five Domains see Mellor, et al. (2020) The 2020 Five Domains Model: including human-animal interactions in assessments of animal welfare; Animals 10 (10) 1870; brief at PubMed (Nat. Libr. Med.) URL https://pubmed.ncbi.nlm.nih.gov/33066335/

[18] Reisner, I (2017) The learning dog: A discussion of training methods; pp. 210 – 226, in Serpell, J (Ed) (2017) The Domestic Dog: Its Evolution, Behavior and Interactions with People: 2nd Ed.: Cambridge University Press; 416pp: provides an excellent view on the training evolution and the debate.

[19] Palman D (2020)  Not Using Force in Police Dog Training; Maine Warden Service; United States Police Canine Association; URL https://uspcak9.memberclicks.net › assets › Resource (PDF); Quote “On the street with a police dog, the dog’s ability to think is an invaluable asset. All of the dog’s natural abilities are put to work for the handler. All the situations that are encountered in real work can never be duplicated during training. I would much prefer to have a thinking dog at my side instead of a preprogrammed machine that has learned to react because he fears punishment”.

[20] Many seem to believe that dogs are just dogs – all just the same beast, which all act the same. I don’t wish to start another YouTube War, but I prefer the Jane Goodall understanding that every dog has its own personality, mind and feelings [e.g. URL: https://www.brainyquote.com/authors/jane-goodall-quotes)]. cf. the mindset of Donaldson, J. (2013) The Culture Clash: A revolutionary new way of understanding the relationship between humans and domestic dogs; Dogwise Pub.; USA; 251pp; Ch1: does not credit dogs with a desire to please or bond with a guardian – although elsewhere (pp. 93, 125/126) proposes praise (and patting) as a reward and encouragement. This view may stem from that authors extensive and great work in rescue shelters with abandoned dogs in desperate need of the basics of survival (also suggested by her insistence that all readers must get their dogs neutered). The remainder of the book is full of really useful, logical information, so keep that open mind.

[21] Here the method of training a dog is referred to, not the rules of the various dog sports, which are set down and must be adhered to so as to achieve an accreditation in that particular endeavor.

[22] Referring here to such elements as puppy toilet training behaviours, acceptable sleeping places for the dog (on/off furniture), etc.

[23] That is, an undesirable behaviour, such as adult dog reactivity; resource guarding; counter surfing; which the guardian desires to alter.

[24] This is not to say that circumstances may never arise when euthanasia is required, but those are rare, sad and special cases when all other means of overcoming the unwanted behaviours have been fully and competently tried, often after a dog attack disaster. See Siracusa, C., Provoost, L. and Reisner I.R. (2017) Dog- and owner-related risk factors for consideration of euthanasia or rehoming before a referral behavioral consultation and for euthanizing or rehoming the dog after the consultation; Journal of Veterinary Behavior; Volume 22, November–December 2017, Pages 46-56; discuss many of the issues and decisions regarding dog euthanasia.

[25] Oxford Dictionary/Oxford Languages: causing strong dislike or disinclination…..relating to or denoting aversion therapy, a type of behaviour therapy designed to make patients give up an undesirable habit by causing them to associate it with an unpleasant effect. "a program of aversive treatment for criminal offenders".

[26] American Veterinary Society of Animal Behavior (2021) Glossary of Terms used in AVSAB Position Statements: www.AVSAB.org: Note that Todd, Z. (2018) Barriers to the adoption of humane dog training methods; Journal of Veterinary Behavior; 25 pp. 28-34, provides a very full summary of all the registered dog associations/clubs globally, as well as veterinary group recommendations, all coming down on the positive reinforcement side of the debate.

[27] Clothier, S (2002) Bones Would Rain from the Sky; Grand Central Publ. NY/Boston; 305pp: at p. 221, disparagingly references Hearne, V (1986) Adam’s Task: Calling the Animals by Name; NY; A. A. Knopf (reprint 2007; Skyhorse Publ.; 288pp), a dog owner who wished to stop her dog from digging holes in her garden. Following the William Koehler methods (Koehler, W. R. (op.cit.)) she had the dog watch her gleefully further dig out the hole, then fill it with water, then hold the dog’s head in the hole under the water - the dog became scared of all holes anywhere – on a walk the dog would skirt suspiciously/fearfully around any depression in the ground. Hearne was delighted with her thorough success.

[28] For an explanation of this game see Wikipedia at URL https://en.wikipedia.org/wiki/Red_hands.

[29] Clothier (op. cit.) puts this very nicely discussing anthropomorphism and allowing that dogs are not totally inhuman: “animals experience pain, grief, love, anger, loyalty and more” – at P.42.

[30] American Psychological Association at URL https://www.apa.org/topics/stress.

[31] Gadbois (op. cit.)

[32] Overmier, J. B. & Seligman, M. E. (1967) Effects of inescapable shock upon subsequent escape and avoidance responding; Journal of Comparative and Physiological Psychology; 63(1):28-33. For an expose of more recent articles ranging from 2002 to 2021 see URL: https://www.sciencedirect.com/topics/medicine-and-dentistry/learned-helplessness#: [Note that a number of experiments described read as unpleasantly as the aversive techniques being discussed]. Maier, S. F. and Seligman, E.P. (2016) Learned Helplessness at Fifty: Insights from Neuroscience; Psychol Rev. 2016 Jul;123(4):349–367; note that "Passivity in response to shock is not learned. It is the default, unlearned response to prolonged aversive events ... " [available at Nat. Libr. Med. URL https://pmc.ncbi.nlm.nih.gov/articles/PMC4920136/ ].

[33] B. F. Skinner (behaviourist advocate and arguably the father of operant conditioning following E. L. Thorndike: both op. cit.) set up this model (further referenced below) and ensuing discussion – which has gone seriously off-track.

[34] Positive Reinforcement; Negative reinforcement; Positive Punishment; Negative Punishment: e.g. “The Preventative Vet” at URL https://www.preventivevet.com/dogs/dog-training-aversives

[35] George Box (1919-2013): see URL: https://asq.org/about-asq/honorary-members/box

[36] Commonly using the Gish-Gallop techniques of accumulating poor arguments to appear like a solid block of evidence; URL https://rationalwiki.org/wiki/Gish_Gallop. As a relevant example discussion see Anderson, E. (2024); URL https://eileenanddogs.com/blog/2024/08/26/.

[37] The David Mech (op. cit.) change in understanding of wolf behaviour is a classic case in point. What a great scientist, ready to correct past misunderstandings of data, concepts and models.

[38] Dweck, C. S. (2006; Rev addition 2017) Mindset: changing the way you think to fulfil your potential; Robinson Little Brown Bk Co., London; 302pp.

[39] Department of Defence (USA) (2019) U.S. Military's Dog Training Handbook: Official Guide for Training Military Working Dogs; US Air Force; Rowman & Littlefield | The Lyons Press; 304pp.

[40] Crippen L (2017) Military Working Dogs: Guardians of the Night; TRADOC; URL https://www.army.mil/article/56965/military_working_dogs_guardians_of_the_night#: Palman (op. cit.).

[41] Haverbeke, A. et al  (2008)  Training methods of military dog handlers and their effects on the team's performances; Applied Animal Behaviour Science Volume 113, Issues 1–3, September 2008, Pages 110-122, describe harsh training methods in the Belgian military environment and observe the “depressed/fearful” (this authors words) body language of the dogs after training – and suggest a more positive reinforcement regime would work more effectively. And see Palman (op. cit.).

[42] Gadbois S (op.cit) expands on this element. c.f. the uncertain outcome of the training method influencing bond/relationship discussed by Ana Catarina Vieira de Castro et al (2019) Carrots versus sticks: The relationship between training methods and dog-owner attachment; Applied Animal Behaviour Science, Volume 219, October 2019, 104831. Also, Ziv, G. (2017). The effects of using aversive training methods in dogs – A review. Journal of Veterinary Behavior, 19, 50-60. [Note that these latter two papers might be somewhat inconclusive in some areas – but the signs are there].

[43] No – not a “get out of jail free card” – an appreciation that in many ways dogs can be more forgiving than humans, but there is always a limit. Again, compare Donaldson (op.cit.) who does not recognise this bonding and desire to please as part of a dog’s nature – having worked with rescue shelter dogs desperate for survival as their priority - not mental enrichment and human bonding.