AVSAB vs Balabanov analysis

Commentary on the AVSAB position statement on humane dog training vs Balabanov panel’s “balanced” dog training philosophy.

Why the Balabanov panel believes AVSAB is wrong.

Introduction

A recent podcast panel constructed by Ivan Balabanov[1] included a distinguished membership of experienced dog trainers and highly educated practitioners[2] who gathered to “dissect the AVSAB position statement on humane dog training[3], reveal its contradictions, and discuss the true science and ethics behind effective dog training”.

The aim of this article is not to be critical of the AVSAB position statement nor the panel individuals and their beliefs. I have followed these panel members closely, watched the videos they make available, bought their books (where available) and hold them in high regard in their field. I would not have composed this article if I totally disregarded their expertise, indeed, I could wish I lived closer to their locales so as to actually meet them and learn more about their dog training methods. The aim is to comment on the views and arguments expounded to determine how well these serve their purpose in “exposing” what they see as the fallacies in the AVSAB statement and the force-free community drive. It’s not about picking sides. In that regard it is important to note that the panel agreed that no trainer they know of uses exclusively aversive methods – they would (I assume) classify themselves as “balanced or inclusive[4] dog trainers”.

Furthermore, I acknowledge that digesting the podcast through repeated views and being able to check the references the panelists quote is a distinct advantage – compared to the panelists “thinking on their feet” and saying what comes to mind. Nonetheless, Balabanov composed this panel of distinguished experts to dissect the AVSAB statement, so why not dissect what the panel states as the “real science and truth”? And the panel attempting to pull apart the precise wording (and the inferences they assume) in the AVSAB statement is also open to analysis.

Podcast intent

At the outset, the brief for the podcast is to “reveal (the AVSAB) contradictions, and discuss the true science and ethics behind effective dog training. Unfortunately, this sets up a strong “confirmation bias” analysis without allowing for interpretation and understanding. Further, stating that the panel will expose the “true science” not only makes the leap that this panel – not the scientists behind AVSAB – are the only ones who can appreciate and comprehend the TRUE science, it is a formula for disaster in terms of fair and proper analysis. Not very “scientific”.

The panelists are also entreated to “stand firm in your beliefs” – again a formula for confirmation bias and preconceived opinions. Perhaps an entreaty to keep an open mind might have been more appropriate?

The claim is made that one of the reasons this podcast was deemed necessary/appropriate is that no one from the force free camp has “stepped up” to debate the issues. Not exactly true, as a number of dog trainers make their views abundantly clear in numerous books and on social media - and a couple of dog trainers have in fact joined the Balabanov podcasts, but been quite severely entrapped in the conversation[5], later characterized by Balabanov as being “polite” (“politeness” is not usually levelled as a criticism) and “skirting around the issues”. I personally know of other eminent practitioners who refuse to expose themselves to what they regard as “unhealthy debate”. Fair enough – their choice.

The force free community are accused of “ideological bias” (not something the panel suffers from, apparently) and attempting to scare dog owners into thinking that they will cause psychological harm to their dogs. The force-free advocates are also lumped together as pushing for the outright ban on certain training tools – prong collars, shock collars, etc.  This is not something AVSAB advocates and it is not helpful to make sweeping statements lumping all those “positive-reinforcement-to-force-free group” of trainers in this one narrow minded basket. My own view is that bans on equipment are not the answer, even though some recent “inventions of tortuous dog training tools” make one want to ban the inventors (see Article on this site – “Should certain dog training devices be banned?”).

The content

The key discussions in this lengthy (2 hours and 40 minute) podcast are challenging to dissect and summarise in an article of reasonable length (so I have opted for unreasonable length) complicated by the uncertainties surrounding some of the research materials referenced by participants and some overlapping discussions and views. That being said, the discussions might be grouped (for the sake of understanding and analysis) into a number of categories (which, by the nature of the podcast and panel to-and-fro, can overlap and blend). These are discussed more or less (but not strictly) in the order they occur during the podcast, with the caveat that certain points may be inadvertently missed – just because there were so very many – no intention of avoiding any position statements.

As a general point, the argument comes up by inference and directly about “how long does it take to train the dog or change a behaviour?” using positive reinforcement versus balanced/aversive techniques - and this is linked to canine, human and even societal welfare. Although acknowledging that we could identify differences in timing, duration and results, I am going to lump this into the “silly question basket” – because there are so very many scenarios; reasons to train a dog, dog natures, dog behavioural problems, etc., etc., etc., that we could find a plethora of diverging answers. If a specialist dog breeder and trainer have a large kennel of dogs from which they plan on either “picking a winner”, qualifying as many dogs as possible for protection sports, or sales to the military/police force, then the timing issue will certainly be at the top of their agenda. However, although the AVSAB position statement is broad brush, one cannot imagine that this small and specialized cohort were a specific target of the statement. If the average dog companions/guardians don’t realise that dog management and training take time and effort, then we are in the wrong debate.

Resisting banning dog training equipment and “dog permission”

The issue of the bans on certain dog training equipment is a simple point of view which comes in two parts. But, again, note that the AVSAB position statement does not indicate a desire to ban certain dog equipment – this is a false inference, if the audience gain that impression[6] - but it is raised as a foundation issue, so let’s discuss it.

The first observation is to all intents and purposes unarguable, viz., in a number of countries/states, there have been blanket bans on certain dog training equipment (the targets are mostly electric shock collars and prong collars) and the aim is to counter this trend and avoid like bans in other countries (in particular in the USA). The podcast does not try to come to grips with why these bans were perpetrated – fair enough, as that could take another lengthy podcast on its own and would all be a bit “after the fact” and so not necessarily very fruitful, apart from illustrative purposes. But then again, it might shed some light on “what went wrong (or right?)” and why that wave of sentiment grew to the extent it did.

The second is that the AVSAB position statement on preferred dog training techniques is characterized as “very harsh and misleading” reflecting an “ideological bias rather than a balanced and evidence-based perspective” by categorically rejecting all aversives and disregarding the complexities of animal behaviour. Now, from the panel’s starting position, this may be a fair statement (although AVSAB equally claim that their position is evidence based) but is it reflective of the underlying intent of AVSAB, which is more along the lines of countering harsh aversives in training (and perhaps stretching to not using aversive techniques – particularly those they specify - as foundation elements in any dog management)?

As part of the sentiment underpinning the AVSAB position statement, AVSAB contend that a dog needs to “give permission” to be trained. The panel puts forward arguments about children who don’t want to go to school – what parent allows this objection to hold sway? Even though parallels between human kid’s and dog’s psychology and behaviour are rarely helpful, nonetheless one can see the point of view - not training a dog because the human worries that they don’t wish to be trained is really a stretch and perhaps a bit of anthropomorphism gone off the rails. At the most, if the dog’s “head is not in the right place”, figure out why that might be (health/distracting environment/etc.), keep the training session short and fun, and take it up tomorrow. And maybe review your training techniques and how you have built the relationship, bond and trust between dog and handler – are you setting up for a “willing dog”?

Definition hiccups

1.     Breeds, guardians and target audience: A subtle, but very important, definitional question is – “What breed of dogs, what type of guardians and target audience (i.e., family companion, protection sports, obedience/agility trial contestants, etc.) do the proponents (not to label them as antagonists) have in the back of their minds?”

The AVSAB document has charming photos of what appear to be “pet dogs” happily interacting with their owners and the text refers to “teaching common training skills as well as to address unwanted behaviours”; “the role of companion animals has evolved, their welfare and the rela­tionships between humans and animals”; “…dogs…(L)learning manners and skills can help animals to co-exist harmoniously with people in the home and in society”. This all implies the focus is on the companion pet dog in the home – although the statements do make a broader sweep on application to all dog training – which is part of the problem, because the broader the statement, the easier it is to contradict. So, one needs to try to see the target and intent – what motivated AVSAB to publish this position statement on humane dog training?

The panel, by comparison, consists of a great team of highly experienced dog experts who own, breed and train (mostly) Belgian Malinois and German Shepherds, with a focus on the protection sports and precision performance style of obedience. This is not to say they do not have “pet dogs” and a strong relationship with their wards/companions, but the focus is different and not subtle.

Why does this “mindset” or approach to the issues matter? Because training methods need to take into account breed characteristics (drive, senses, “personalities” – “strength of character”) and to state that training methods for all dogs should be the same is just a bridge too far. No – it doesn’t mean that one group of dogs needs aversives and the other group not, it simply means that making sweeping, all inclusive, statements in the context is on thin ice – more consideration needs to be given to the dog, the purpose and the particular training challenge.

2.     The next foremost glaring definitional absence is a clear consensus on what the panel regard as “aversives” and in what situation of application. And therein lies a major issue with the panel discussion, in that whilst AVSAB characterizes aversives as “any use of training tools that involve pain (choke chains, prong collars, or electronic shock collars), intimidation (squirt bottles, shaker noise cans, compressed air cans, shouting, staring, or forceful manipulation such as “alpha rolls” or “dominance downs”), physical correction techniques (leash jerking, physical force), or flooding (“exposure”)”, the panel does not clearly counter with a list of what they regard as “behaviour modification or training-essential aversives” and, in fact, each panel member actually appears to agree with the AVSAB statement, in that they disagree with the harsh punishments listed, at the least as potentially badly used by incompetent trainers.

3.     A further missing foundation in the discussions is that it is never clear whether simple training (“obedience/agility/tricks/etc.) is being discussed, or undesirable behaviour modification[7], or indeed the “tougher” training regimes of the protection sports or military dog training. Further, as noted above, the issue of training a Cavalier King Charles Spaniel or a Belgian Malinois across all of these training/behaviour modification scenarios – can we really discuss training methods for such unlike breeds , in the same breath? Again, the sweeping statement challenge. These camps are so very largely separated that the discussions and arguments find themselves at complete cross-purposes, racing along in opposite directions on the two opposing sides of a divided highway. Another way to look at this problem is that blanket statements are by their nature “absolutist and allow no qualification” and can so easily be taken out of context, misunderstood or countered by a single example to the contrary. In this respect, the AVSAB dilemma and pitfall is that to remain brief and to get their point across, they have limited themselves to blanket statements. It doesn’t work as well as they might have hoped.

4.     The “trust vs welfare” definitions. Actually, both of these are reasonably well understood terms in common parlance[8] (although the claim is made that “trust” is not a “scientifically” defined term) and both have a valid place in the dog training and psychology lexicons. The panel goes into complex definitions of welfare, questioning if AVSAB might mean the presence or absence of stress related to an environmental challenge or an adverse emotional response in the dog (usually acute). However, it is argued by the panel that these do not necessarily define the welfare state, but rather what is key is the absence of predictability and control by the subject (dog) or can the dog successfully cope with the situation (referring to the applied/implied aversive) by foreseeing consequences and controlling its own behaviour to avoid unpleasant outcomes. This line of thought expands to consider not just the immediate (observable) stress a dog may feel, but considers the welfare surrounding the dog not being trained to perform or cope. In terms of trust, again the point is made by the panel that the dog has a degree of “agency” in the training regime (and life in general) if it can both predict and control what will happen if it takes a certain action or acts out a certain behaviour – how reliably will the result occur and is it a “fair outcome” (presumably according to the dog itself)?

It would be fair to guess that this discussion goes beyond what AVSAB had in mind regarding a dog’s welfare – they may have been simply thinking “wellbeing” in a general sense. Perhaps “quality of life” over an extended period, as a panellist points out, would have been a better place to start, rather than focussing on the moments surrounding aversive training – but that is quite hard.  Again – the challenge of a focussed and brief position statement attempting to cover all the bases.

The panel extend this welfare definition by making the fair point that “welfare” needs to consider both the dog and the guardian in terms of what they can and cannot do in their day-to-day lives and enjoying enrichment activities together. We are told that hundreds/thousands of dogs are trained with aversives and they have great canine-human relations. Not a very technical argument (and see Footnote 11 below) but I would make the point, as I have elsewhere, 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[9].

5.     The definition of “stress” and how it is applied. There is the stress of learning Obedience, Agility, Protection Sports, Tracking, etc., disciplines which all exercise a degree of stress placed on the mind and body. This needs to be distinguished from the uncontrolled/unpredicted stress of aversives which (may) build to “distress”- that is, without time or ability to recover equanimity[10].

6.     Finally, constant, passionate reference to “science”, “evidence-based” and even “personal experience”[11] is entered into by both sides of the debate. Both sides seem able to adequately prove their points – and how is this helping us advance the dog training methodologies we all strive for? As an aside, the point is made that there are “absolutes” in science – fundamental scientific laws that will never break. Yep – that’s what the Flat Earthers said and why that historical polymath, Copernicus, was not well received by the scientific cohort of the time. Scientific conclusions are made in the context of empirical evidence, observations, data records, observer bias, unknown contingencies and the state of knowledge (and social mores) extant.

Aversive training unnecessary vs “If there was a better way….”

The AVSAB statement includes the contention that (paraphrasing) “There is no evidence that aversive training is necessary (nor more effective than reward-based methods) for dog training or behavior modification (in any context)” – which the panel claims is “patently false” and damaging to dog welfare (leading to relinquishment to shelters/euthanasia as the end result) specifically quoting a recent study on efficacy of electric collars in managing/halting predatory behaviour[12].

On managing predatory behaviour, the panel did seem to agree that the quoted research (Footnote 12) does not relate well to what happens in “real life” electric collar training and use – a valid point to be remembered when quoting “scientific research” to prove a point about any dog training methods. Further, as noted in another Article on this website (“Should certain dog training devices be banned?”) when a dog off-leash has the predatory drive, there is some evidence to indicate that even a high-level pulse on an e-collar will not divert the dog from its target in this situation[13]. Nor will the offer of a high-value treat in that moment – when the dog is also deaf to cues from the guardian[14]. The predatory instinct under extreme provocation is incredibly strong – a primal drive[15].

AVSAB back their position by noting, for dogs with a history of off-leash behavior problems, that research[16] found no difference in the pro­portion of disobeyed cues between dogs trained with electronic shock collars by manufacturer-nominated trainers com­pared with reward-based training. Dogs trained with reward-based methods in this study had a shorter delay before respond­ing than the group trained with electronic shock collars.

Perhaps the point remains unresolved when it comes to stopping predatory behaviour using a shock collar – but stating that the AVSAB position is “patently false” is stretching it a bit.

The point is made by the panel that “if there was a better way of dog training” (than “balanced training or using aversives”) then the panel members would universally follow that way. A convoluted argument which bites its own tail, as the panel all agreed that on the one hand, their own attitudes towards what they regard as acceptable in dog training shifted enormously under the influence of the positive reinforcement revolution and what they referred to as the “signaling evolution” (the ability to develop communication systems with animals) - so the panel agreed that they are the happy recipients of the best dog training of all time; and on the other hand, they always start training using positive reinforcement because it works so well.

So – they found a better way? But then find they need to resort to aversives because….? To think positively, the panel did say that using force-free/positive reinforcement is not the only requirement, but they did say it was pretty fantastic that this revolution in thinking occurred and strongly influenced them. This argument is not grappled with, so a cynic might conclude that aversives are the frustrated trainer’s haven, or a more generous mind might conclude that after basic learning the application of some (mild/undefined) aversives plays an important (necessary?) role in the dog’s training. Either way, they agree that they “did find a better way”, at least in part.

Somewhat predictably, given the panelists invested interests in particular dog breeds and dog sports, the argument is presented that the ultimate proof of the need for and value of using aversives in training comes from the successful winners in protection sports competitions. Add that to the twin baskets of silly arguments and sweeping statements. Firstly, the training methods of all global competitors cannot possibly be known, nor can their definition and “severity” of aversives – if they use them at all. Secondly, this is classic “talking your own book”- viz. a sort of combination of patting yourself on the back and stating an argument which benefits you because of your personal involvement or vested interest. Although I enjoy watching videos of well-trained military, police and protection sport dogs and admire the athleticism, precision and (mostly) control, not everyone in the dog world holds winners of dog protection sports in high regard – there are other equally valid measures of good dog trainers and good dog companions. I also – to make a leap – have a sense that AVSAB were not thinking along these lines or targeting this group of dog trainers in their position statement.

Let’s be absolutely clear – it is the view of this author that all the panel members are outstanding dog trainers in their chosen field and none of them resort to animal abuse. They clearly state that none of them endorse generating fear, intimidation or pain in a dog as a process of training. What they do endorse is the use of “aversive control” which is “effective and efficient”. Although we once again trip over definitions and broad statements here, these practitioners, I believe, simply could not do what they do, or come across in the genuine and passionate manner that they do (albeit I disagree with some of the arguments they put and some reliance on prong and shock collars[17]) unless they cared deeply about dogs and how to be a great dog trainer.

A point is made – loosely relating positive reinforcement to aversive methods - that the theoretically force-free teaching technique of “free shaping” is actually driven by frustrating the dog through denying the reward until it takes the next step in the desired action or activity (i.e. applying aversive coercion). A somewhat strange training philosophy. One would rather characterize the shaping technique as making the “next step in the sequence” obvious to the dog, for example by moving the reward (the “lure”) slightly further away or in a slightly different position than it was in the previous iterations[18]. This is not exactly “errorless learning” (more a part of the “chaining” technique) – but we will get to that.  

Cortisol levels as indicator of stress

The AVSAB statement refers to cortisol levels[19] as being found to be higher in dogs trained using aversive methods (when measured right after training[20]) indicating stress and harm to the dog has occurred through the use of aversives. The panel quite rightly point out that positive stimulation also raises cortisol, so taking this measure alone might add to other information/observations, but it is not a standalone indicator[21]. Paraphrasing the panels thoughts on this: “(what is) … important is essentially the rate of recovery of a corticosteroid spike, not the absolute magnitude of the corticosteroid spike, so levels of circulating cortical steroids are worthwhile and important indicators to look at, but it is far from as simple as looking at cortical steroid levels at a single point in time - that is why most researchers also use converging measures of behaviour….”.

The whole subject of cortisol in dogs is far more complex than covered by AVSAB or the panel (naturally – as the focus was purely on aversives and stress causation). There is an overlay on cortisol levels of its impact by environmental factors, human interaction and even auditory input (such as music) – and the impacts are not just on behavioural matters, but also disease[22].

A further interesting reflection on this issue – not described by AVSAB nor the panel – is that Schobert, et al. (2017)[23] found that coping with stress in a human-canine dyad can be mutually influenced by the human-canine bond and individual states of “being” of the pair - that is, “emotional contagion” or “mutual emotional support”[24]. The fascinating findings of this study are beyond the purview of this paper, but it does raise the question as to the possible influences of the emotional state of the trainer (and, interestingly, their gender), the human-canine relationship/bond and the manner in which aversives may (or may not) be applied and therefore play a role in obedience training or managing unwanted canine behaviour.

Cognitive bias

In the context referenced here, cognitive bias refers to a choice that a dog might tend to make based on how optimistic or pessimistic it feels – and that in turn is impacted by its treatment during training. So, it is not unrelated to welfare and trust factors (refer again Footnote 8 for definitions of these terms). The “cognitive bias” paradigm is based on findings that affective state can influence cognitive processes in humans. For example, people in negative affective states judge ambiguous stimuli negatively (are relatively ‘pessimistic’) compared to happier individuals[25].

Hence, AVSAB describes dogs trained with aversive methods being more “pessimistic” (as well as having strained bonds with the guardian) on average compared to dogs trained using positive reward-based methods[26]. The panel tends to discount this view by implying that it is not well proved science[27] and at the end of the day if a dog has to go to the vet and is pessimistic about the outcome – so what? It still has to go to the vet – so practicality overrides how the animal is feeling psychologically. This seems to make sense in the vet visit example, but skates past the AVSAB point, which is that training outcomes (and welfare state) are suggested to be better if a dog is treated so that it develops optimism rather than pessimism. And yes, there are “scientific” ways to define and test optimism versus pessimism.

Stress prepares a dog for life’s eventualities

The argument is made in the panel discussion that stress in a dog’s life is essential to prepare for life’s eventualities, either because “not everything in life goes right all the time”, or to prepare for unavoidable events, such as coping with the stress of going to the vet. A dog which has not experienced stress (through the application of aversives in its training, one assumes) will be less able/unable to cope with future stressful situations. That is, we impoverish dogs if we protect them from all stress – by sheltering dogs from stress their ability to cope with future stressful situations is depleted, increasing the intensity of responses to stress and decreasing the ability to recover. In short, the panel suggests that using aversive training methods builds resilience in the dog (meaning mental stamina built through the exercise of handling aversive training techniques?).

Firstly, to ensure proper context, the AVSAB statement doesn’t suggest “sheltering” a dog from life, nor that a dog experiences no stress whatsoever, but rather holds that “…(T)he techniques used to teach these manners and skills can strongly affect an animal’s future behavior and emotional wellbeing” and goes on to claim that “… studies have shown an associ­ation between the use of aversive training methods and long-term behavior problems including aggressive behavior towards people and other dogs, and anxiety-related behaviours such as avoidance and excitabil­ity. Survey studies cannot differentiate between causation and correlation, so possible explanations for this association include: 1) aversive training methods directly cause or contribute to the de­velopment of problem behaviours; or 2) owners of dogs with problem behaviours are more likely to use aversive training tools”.

Now, going back to “stress in a dog’s life is essential to prepare for life’s eventualities” as suggested by the panel, this argument suffers badly on two fronts – remembering that AVSAB does not say a dog should experience no stress of any kind whatsoever (on what utopian planet could that occur for any living being?) but rather consider application of aversives in dog training/behaviour management. So, what is required to compare these opposing positions?

Firstly, define the source, type, repetition and duration of stress – not all stress is the same. Experiencing the stress of learning a new action or routine is very different from experiencing the stress induced by aversive treatment in training. Either way, these stresses are not in any way going to prepare a dog for the drama it feels in a veterinary visit, where past experience tells it bad stuff happens and the “alarm pheromones” are working overtime. Is the suggestion really that application of training aversives can counter the impact of pheromones? Every vet would love to learn that trick!

Secondly, in what context can experiencing an aversive or any stress on the training field possibly prepare a dog for a future threatening, frightening or stressful situation? Well – there is one relevant context that comes to mind - and that is in the training of military and police dogs learning to combat assailant threats and stressful predicaments (like sky diving into a war zone strapped to their handler). But this is not achieved through applying aversives, it is achieved through repeated play acting of the scenario build-up (“chaining”) with a positive outcome (viz., the dog is tested and “pushed”, encouraged and rewarded, but not severely damaged nor does it need to be punished with aversive actions).

The panel makes reference to a study[28] which tested the hypothesis that although “unpredictable chronic[29] stress” resulted in enhanced depressive and anxiety like behaviours and memory dysfunction of rats in the experiments, by comparison the effects of “predictable chronic mild stress” – such as routine day-to-day life stress - had impacts on mood and memory, effectively enhancing brain function[30]. That is, “predictable chronic mild stress” overall delivers potentially beneficial outcomes in terms of both mood and cognitive function. Now, in considering this “knock me down with a feather proof” of the benefits of stress –  remember (a) it is one set of experiments on rats, not dogs, being stressed by forced swims and maze tests, (b) the amount of stress that would be relevant (if at all) in the dog training context is completely unclear and (c) the correlation between these stresses and using aversives in dog training is undefined – one has to draw a long bow and do a very broad jump to any conclusions regarding dog training, as tempting as it is to correlate the rat’s long swims and maze work in the quoted experiments with dog training such as Agility. But in any event, that does not aid the “use of aversives is beneficial” argument, nor confound the “umbrella intent” of the AVSAB position statement – except to counter the totally implied/imagined/ stress-free existence.  Again, defining “aversives” and when and how they are applied and how this relates to the stress described[31] are key elements completely absent from this discussion.

Taking a leap of faith, let’s assume that in suggesting that stress builds “life preparedness”, the panel is actually referring to building self-confidence and resilience in a dog (this could, of course, not be what they meant at all). Self-confidence and resilience are closely related to “willingness” of a dog in learning, which in turn is related (this author’s contention) to the relationship, bond and trust between canine and human, as well as relating to the dog’s nature and, indeed, the efficacy of the classical and operant conditioning experienced (and avoid thinking that operant conditioning calls for all four quadrants – it doesn’t)[32]. These elements are all built through “experiential learning” – what some refer to as socializing a pup, not just to people and other canines, but to all sorts of experiences, environments, physical and mental exercise regimes, walking surfaces, visual and auditory impacts, and so on.

Dog’s play involves aversives: aversives are evidenced by human comparisons and history

The suggestion is made by the panel that when dogs play, they bite at each other and invite biting back – and this is aversive – and so aversives are fun and part of what dogs do. Really? Not even a sensible argument and the linkage to using aversives in dog training is completely obscure. Dogs use their mouths in many forms of communication – with conspecifics and with other species – and they moderate their biting behaviour according to the situation. In a litter of puppies, if one pup bites too hard, the other lets it know in no uncertain terms and will not continue the play.

This argument is somehow extended (in a convoluted way) to reference using aversive actions to correct humans who self-harm to prevent a recurrence. This is presented as evidence that to state (by AVSAB) that “aversives play no role in behaviour modification” is completely wrong – because they do in the most extreme human dysfunctional behaviour. Hang on folks – I thought we were discussing dog behaviours? To compound this misguided argument, it is presented that “historically, aversives play a role in behaviour modification” – really – are we thinking historically about the Spanish Inquisition here as the historical evidence to tell us how to behave today? Not even remotely relevant. Perhaps if a parallel was drawn between a severely aggressive dog (or one that self-harms out of anger or frustration?) then maybe we could see some loose correlation with using aversives in training – but the panel fails in this regard and, in any case, it’s doubtful.

As so often before, the very tenuous connection is drawn between humans engaging in risk taking sports activities and the act of developing expertise (through sports training) and thereby managing the risk (and aversive contingencies involved) is reinforcing. The leap is that it is then stated that dogs experience the same positive sensation when exposed to aversive contingencies. This is confusion of risk, aversives, stress, and learned behaviour all bundled into one ball of knots. A dog learning Agility, Hoops, Water Rescue, etc., all involve an element of risk contingencies and directed effort (don’t slip off the Agility ramp, don’t crash into the hoop or jump, don’t drown, etc.). But what does this have to do with the justification of applying undefined aversives in the training methodology? Suggestion – nothing - and the proponents need to realise how silly this makes their arguments appear (the panel mostly looked a bit non-plussed when this argument was presented – but – were too polite to disagree?).

Avoidance conditioning/Aversive instrumental conditioning

There is then a rather detailed quoting of research (targeting the field of human psychology, not dog training) on the subject of “avoidance conditioning” (“aversive instrumental conditioning”, also referred to as “fear learning” and elsewhere in the podcast loosely as “apprehended contingencies”) and responses[33] - targeted at justifying the use of aversives – or even requiring that usage(?) – on the basis that “fear” does not play such a big part as previously imagined in applying “aversive contingencies”. This discussion raises the debate to a somewhat esoteric level, probably above what AVSAB targeted and the average pet dog owner wants to get involved in, but having been thrown into the discussion it needs to be addressed. The end point of the argument is that a “sophisticated dog trainer with high level technical skills” is able to apply “well conducted avoidance conditioning” – avoiding the terminology around fear by relabeling it “discrete and limited periods of stress causing apprehension but not fear (redefining “apprehension” to mean something like “understand the consequences or the contingencies that are both predictable and controllable” as constructed by their savvy trainer). Is this clarification or obfuscation?

Avoidance conditioning is explained in the panel through a parallel between a man wearing sandals on his feet so he can walk across a stoney car park – he predicts the challenge and avoids the pain (which has been conditioned - one assumes - because one time he tried walking across the car park barefoot). The man does not fear the sandals, nor the carpark now. The similarity is drawn with a dog wearing an electric shock collar and learning that it can control/avoid an unpleasant (painful) outcome if it performs as commanded by the handler. Now, the human has complete control (“agency”) in whether or not to walk across the carpark and whether or not to wear sandals – so little wonder he doesn’t fear the sandals or the carpark. Is it really the exact same for a dog wearing a shock collar (unless, of course, it chose to wear it)? Sorry – there may be a vague point to appreciate here, but that’s it - the human-canine parallel fails. Note that the negative aspects of the application of avoidance conditioning are that when the triggers – the “aversive instrumental conditioning” - “are activated in a recurring, excessive, or prolonged way that goes beyond the requirements of the situation”, then the fear responses can be maladaptive (refer again Footnote 33). So, simplifying all this to say, “Hey – it’s not as bad as previously thought – if applied by a highly skilled trainer” is a possibility – but is it really a leap of faith?

Notwithstanding, this whole subject (based on human psychology/behaviour combined with animals research – mainly rats, but some reference to past experiments using dogs) is a heck of an interesting concept, because it just jumped to a conclusion about a whole raft of animal psychology and tells us that, if it can be applied to dogs, then just like humans, dogs can reflect on the past and envisage future scenarios, not a popular line of reasoning, but one for which there is at least some support[34]. Dogs do learn to predict the future events – such as when the guardian picks up the leash or the car keys the dog anticipates an outing – but this is simple Classical Conditioning. However, this “sandals in the car park compared to electric collar use” goes way beyond classical and operant conditioning – and it would be kind of nice if it was true and dogs really thought like that. Importantly, realise that this human comparison and proposal leaps outside of the “Behaviourism” psychological model into “Cognitivism”, “Constructivism” and “Experiential” psychological models. Interesting hypothesis – would love to see this researched – could start another word-war between Behaviourists and other branches of psychology.

A further branch of this “apprehended contingencies” is explained by panel members riding dirt bikes, rock climbing, learning Ju Jitzu and working hard to improve their skills, because if they don’t, they will suffer a consequence. The jump here is that humans are said, therefore, to actually be seeking out or “chasing” those aversive consequences so they can prove to themselves that they manage the outcomes – or avoid the worst of them – by developing a skill set. With success in these ventures, that outcome management or avoidance of hurtful consequences in itself becomes highly reinforcing – for the humans. Dogs are thought to share this same phenomenon of “aversive avoidance” being highly reinforcing. Wow! Maybe? Interesting take on self-directed behaviours and aversives/punishment equaling reinforcement. So – Thorndike and Skinner were just looking at it wrong and Pavlov didn’t go far enough – but perhaps this tells us that Most and Koehler actually had it right all along?[35]. I don't think so.

The argument inevitably returns to a contention that all animals (all-inclusive, but focus on humans and hence dogs) have the “approach and avoid” innate sense baked into their DNA, which somehow extends to argue that therefore aversive application in dog training is fundamental. Animals are designed to resolve problems, apparently including punishment and aversive scenarios. So – it is argued – generally using aversives in well-structured dog training is fine - a trainer would have to deliberately go out of their way to deliver inappropriate/forceful/fear inducing aversives in bad training scenarios. Really – it’s not just about non-contingent punishment; it is too broad a spectrum to pass it off in this manner. This is not adding to the resolution of the debate over the veracity of the AVSAB position statement. Again, drawing parallels between human thought and animal thought is fraught with difficulties.

Errorless learning

“Errorless learning”, as an alternative to standard trial-and-error learning, relies on the tested hypothesis that errorless learning minimizes errors by providing few initial opportunities for incorrect responses to the incorrect stimuli[36]. Note that this discussion is an extension (perhaps only “by inference”, which is fine) but not something the AVSAB statement suggested should be applied to dog training – but the panel raised it, so it is worth discussing.  Now, although this is a complex field, a basic outcome of research in this field has been that the “errorless learning dogs” were more likely to meet the criteria for learning the required discrimination, showed faster speeds of acquisition and displayed fewer frustration-related behaviours, compared to the trial-and-error group. The hypothesis is that dogs might be trained more effectively and with better welfare during training, suggesting errorless learning could be more greatly utilized for canine training and research[37].

The panel appear to discount errorless learning as either a myth or a fad (related in some circles to the failings of the teacher rather than the student[38]) and in any case missing out the important step of putting a dog under some element of stress (presumably through failed responses and the consequent application of aversives) be that for some future-proofing benefit or exhilaration. Once a dog has learned a particular process or objective, can “trial and error” play a part? Certainly – but that still does not require aversive techniques.

Pavlovian fallout

A companion line of thought - or a feeder to errorless learning - is covered under the ‘Pavlovian fallout” concept (perhaps inferred but not directly named by AVSAB). According to the work of Sidman[39], (paraphrasing) “we use coercion almost exclusively to control each other; many find it hard to imagine any other way. The author asks, does the death penalty deter potential murderers? Is harsh retaliation the answer to the discipline problem in our schools? Do the standard coercive practices work? – in law enforcement, behavior therapy, education, the family, business, the armed forces, diplomacy. Behavior analysis has shown that they do not work. Coercion is in the long run self-defeating. Punishment eventually proves counterproductive”.

The contention is that punishment does not suppress the targeted undesired behaviour but rather makes the dog “apprehensive” of the human-canine relationship. This is “Pavlovian fallout”- self-defeating in terms of the targeted outcome and disengages the dog from the training regime. The panel argues that in fact the Sidman’s (and Skinner’s) contention that therefore punishment does not work is wrong – but do admit that if the aversive/punishment administered is “powerful” (undefined) it tends to be associated, via “Pavlovian mechanisms”, with all the stimuli in that situation – and that is fallout impacting the dog’s willingness to interact with the trainer (and perhaps that particular training environment) and perform as cued/commanded. So, in administering an aversive, all the elements of timing, relevancy and appropriateness (collectively labelled as “contingency and contiguity”) so that the dog forms a clear linkage between the action and the punishment are imperative. A good argument, well explained – but well outside the AVSAB arena.

Moreover, there is a suggestion that when this “fallout” does occur, it fades over time and the responses that are not selectively punished, recover to normal levels - and the specifically selected punished “Target Behavior” remains suppressed. That is, the panel argues, fallout is not inevitable, it is not permanent, it can be managed and it can be minimized. The panel argument goes on to contend that “if you're going to do technically strong dog training using aversives you have to know about it and you have to know how to mitigate it, how to minimize it and how to restrict its time course so it's just a little phase the animal goes through until he develops Mastery” (effectively the “positive psychology” and “resilience theory” or “optimism” of Seligman[40]). Again, probably not thinking of the family pet Chihuahua here.

Right or wrong, as a broad statement applying to all dogs, this delivers the desired excuse to those leaning towards aversive training methods, effectively countering the generalities of the AVSAB position statement by requiring a higher level of sophistication, expertise and sensitivity in the dog trainer and a responsive (strong temperament?) dog. In other words (quoting the panel) “the dog is properly handled and dealt with in the course of training featuring aversives and the period during which the animal looks like it might be vulnerable, or its welfare might be in question, can be rendered relatively innocuous and short in time span to be replaced by Mastery, which is associated with prevailing positive affect, meaning the animal looks like he enjoys training - he looks like …. he's controlling things, he has agency.… he has the power to dictate his immediate circumstances within the contingencies that we dictate as his trainers who are responsible to him and to society and to other animals to globally take care of everyone's welfare by optimizing the behavior of that animal”.

Now, that is all pretty convincing – at least for the sophisticated trainer – but still begs the questions as to (a) were the aversives actually required in that undefined predicament and were they beneficial to the training outcome and (b) given that the trainer needs a high level of skill, does it work for every dog and every guardian/trainer, universally - and if that is questionable, doesn’t the AVSAB position still have a degree of validity and value – at least for most dog companions/guardians and would-be dog trainers?

Summary and conclusions

As a summary issue, the panel ask that if we are to believe the AVSAB statement that there is no role for aversive training in behavior modification plans, does that mean that you cannot really even say “Hey, don't do that” and don't even use the word “No” and have a consequence for doing it? I would agree that this is an extreme view from the totally force-free community (discussed in another Article on this website “Reinforcement vs punishment: A dog trainer’s perspective”) but I don’t see that extreme view in the AVSAB position. Projecting (or channelling) the extreme force-free fringe onto AVSAB is unfair, not helpful and possibly even misleading.  

I also agree with the panel that AVSAB have a position which is appealing to a large number (majority?) of dog guardians, in that people generally aspire to “being nice” and having a well-trained dog without causing fear, distress or harm to the dog. That’s just how it is – so, the “balanced” dog training community, if they are to enter this debate, need to be a bit smarter in “targeting” their arguments. At the moment, the balanced or “pro-aversives” tend to come across pushing the views that punishment and aversives are absolutely required and anyway, that’s not all bad, because that’s how life works, we face aversive events and learn to deal with them and dogs are just like people - and even aversives are really reinforcement when they are overcome. So, the punishment/aversive cartoon looms large in people’s minds from the outset. Surprising the two side can’t seem to align? Not really.

Panel members agree that they have encountered people that come to them for training with their dog that has been through several “all positive” trainers and they feel ashamed and guilty because they are told they are not achieving their training goals because they are not following the protocol right - not that the protocol is at fault. I agree that this can obviously happen, but equally I have had numerous pet dog companions (guardians) seeking help because a previous trainer was aggressive and forceful (even violent) with their dog, which set them back months in their training goals and in a number of cases even put them off training altogether[41]. So – does this prove that either positive reinforcement or balanced trainers are all idiots giving bad advice, or could we just agree that there are two sides to this coin? There are good and bad dog trainers giving advice and the dog guardian needs to make a judgement as to what seems right, fair and appropriate for their dog. Not easy, of course, because just like when a person seeks help from any “expert specialist” in any field, they are relying on the “knowledgeable advice” they have been looking for. Either way, the debate is not won or lost by noting that there are misguided extreme dog trainers out there.

To reiterate, the aim of this article is not to be critical of AVSAB nor the individual participants on the Balabanov panel and their beliefs. The aim is to comment on the views and arguments expounded to determine how well these serve their purpose in “exposing” what they see as the fallacies in the AVSAB statement and what they characterize – rightly or wrongly – as the force-free community drive.

So – who is right – who won this “debate”, AVSAB or the Balabanov panel? Neither. Not exactly zero marks to each side, but there is a truism that “a person convinced against their will is a person not convinced at all”. Do the panel arguments presented actually undermine the AVSAB position? No – not least because the two sides are so much at cross purposes, absent agreement on definitions and intent (the “arena for the battle”).

Then, is there a way forward? Well yes, but the debate should be about attempting to see a way to overcome this divide in dog training philosophies, which is doing little to further the ambition of all dog trainers, viz., to better understand the options, protocols and constraints in dog training and ultimately to become better at the craft – and perhaps to share that knowledge with others, devoid of emotion, spurious arguments and trying to beat the opposition to death with “science” quotes. As the panel agrees, research findings under controlled conditions do not necessarily reflect how dogs are actually trained in the field, day to day.

So, what is required is starting with an open mind rather than a firmly held belief continually relying on preconceived ideas and confirmation bias. It will also require, as discussed, an agreement on key definitions, in particular what are the aversives, categories of stress, timing factors, the training goal (obedience/manners/ sports/undesirable behaviour modification/etc.) and which dog types (breeds/personalities/ drives/etc.) an argument applies to. Oh yes, and avoiding silly comparisons between humans and canines (whether it is playing a risk-oriented sport or deciding to speed on the freeway) to prove a point and staying away from “science proves that”. Consider the research and the personal observations by all means (they are important) and sure, use a human analogy to illustrate what is meant – but don’t think that analogy will prove how dogs think and react – it will not. Be aware that there is a plethora of research and thousands of personal observations of the complex variability of the natures and behaviours of dogs across all canine training fields, from pet dog manners training to highly trained dogs in protection sports and military/law enforcement roles.

Perhaps the way to consolidate and properly utilise the dog training/behaviour experience and research might be – before we dive in to citing evidence and expressing opinions about the rules/guidelines/codes/principles of dog training - to start with categorizing the debate according to the “key definitions” described. Easy? Nope! A veritable Rubik’s CubeTM of options and opinions. But thinking about these parameters might just help the debating sides figure out what they could agree upon and what they could learn from each other?

I do still hold out some hope……….

[If in any of the foregoing I have “put words in the panel’s mouths”, or misinterpreted the AVSAB position statement, that is inadvertent and would have arisen by my trying as far as possible to appreciate the arguments put forward. There is no intentional misinterpretation or misrepresentation of the panel’s intent – but if that has occurred, then the fault is mine in misunderstanding].

[1] Ivan Balabanov Podcast Episode 52; December 2024; YouTube https://youtu.be/Qy0Mz9HCEUU?si=ErnYieFxmAJuOG53  

[2] Stephanie Cottee, an animal behavior and welfare scientist, researcher, and co-owner of Black Kraken K9; Dr. Stewart Hilliard, a behavioral neuroscientist with a PhD and 45 years of experience in training and breeding working (and military) dogs; Michael Ellis, a well-known trainer with 45 years of experience - founder of the Michael Ellis School for Dog Trainers, and a breeder of exceptional Belgian Malinois; Balabanov’s credentials and success are well know from his YouTube channel.

[3] AVSAB -The American Veterinary Society of Animal Behavior; www.AVSAB.org; Position Statement on Humane Dog Training; Evidence (downloadable PDF) supports the use of reward-based methods for all canine training; AVSAB is “a professional organization that unites veterinarians and research professionals who share an interest in both understanding animal behavior and treating behavior problems that affect the welfare of animals and the people who care for them". The statement is supported by the American Veterinary Medical Association; URL https://www.avma.org/javma-news/2021-11-01/veterinary-behaviorists-no-role-aversive-dog-training-practices 

[4] This renaming or rephrasing of key words is common in the dog training debate – just as choke chain collars become check chain collars, prong collars become pinch collars, electric shock collars become stim collars – now use of aversives or balanced training or aversive intervention become inclusive training or inclusive intervention? It does have a nice “woke” ring to it (?).

[5] Such podcasts tend to be one-sided just because the instigator has an agenda and pre-prepared questions in mind – so the thread of the conversation can easily (unfortunately) trap the respondent.

[6] Early in the podcast introduction, reference is made to “it's very critical for the future of our industry here in the States not to bow to all those extreme, all positive ideology forces, because we know what happened in Europe and that that the chances to happen here is just as good if we don't maintain some reasonable defence to that”.

[7] The panel makes reference to “medication” in behaviour management – although not expanded upon, one assumes that the panel might think that training methods – including aversives – would be a better option(?). If that is the case, I would agree – for many cases, just resorting to medication as the first port of call is not the answer, but again, beware the sweeping statement. I have worked alongside vet behaviourists on dogs that are so strung out and constantly triggered to a fearful and reactive state, they really need help beyond just training to assist them to “carry the mental load”. Medication in these cases has opened the door to applying sensible classical and operant conditioning-based training to “re-program” these dogs and give them a better life.

[8] Oxford languages definitions: Trust - firm belief in the reliability, truth, or ability of someone or something: "Relations have to be built on trust". Welfare - the health, happiness, and fortunes of a person or group.

[9] See Articles on this website “Reinforcement vs punishment: A dog trainer’s perspective” and “Zen and the Art of Companion Dog Training”.

[10] This point is made in another Article on this website “Reinforcement vs punishment: A dog trainer’s perspective”, referencing the argument of the guy in the gym experiencing mind and body stress – not distress.

[11] This “personal experience” argument reminds one of the red herring that “I was brought up that way, and I turned out alright!”

[12] Johnson, A. C. and Wynne, C. D. L. (2024) Comparison of the Efficacy of Different Training Methods in Stopping Chasing Behaviour in Dogs; Animals; 14, 2632; URL https://doi.org/10.3390/ani14182632

[13] “… Increased arousal – increased arousal levels increase your dog’s pain tolerance (this is a strong survival instinct). This explains why a dog who has severe silent pain can still chase after a ball or compete in sports such as agility or flyball….” from Forest Canine: Pain in dogs - The Harsh Reality; URL https://www.forestcanine.co.uk/post/silent-pain-in-dogs-the-harsh-reality 

[14] “When dogs become excited, their adrenaline levels rise, leading to a state of heightened arousal. In this state, their attention span decreases, making it challenging for them to focus on commands. Additionally, excitement can override their training, causing them to prioritise the source of their excitement over obedience….” from Paws, Claws and Tails: URL https://pawsclawstails.com.au/unlocking-the-mystery-why-does-my-dog-ignore-commands-when-excited/ .

[15] Finally, talk to some farmers or hunters about how their stock/cattle/hunting dogs chase prey (like deer, kangaroos or wild boar) through several barbed-wire fences, get severely damaged (by brambles, the wire or the boar they catch), but don’t stop the chase, to the point where some might make it home at the end of the day, but don’t survive the damage.

[16] China, L. et al (2020) Efficacy of Dog Training with and without Remote Electronic Collars vs. a Focus on Positive Reinforcement. Front Vet Sci. 2020; 7:508.

[17] One does worry, given the context of other arguments the panel presents, that there is an element of “the end justifies the means” playing a part in the choice to use such tools.

[18] Burman, O. (2008) Sensitivity to reward loss as an indicator of animal emotion and welfare; J Biol. Letters; V. 4; pp.330-333; URL https://api.semanticscholar.org/CorpusID:1941970  delivers an interesting perspective in that the research found that rats that were deprived of an enriched life in their experiment suffered a prolonged response to the loss of expected rewards more severely than rats which had a higher standard of welfare and enrichment, which perhaps links to a following point on “cognitive bias”.

[19] Cortisol is one of the vital steroid hormones (glucocorticoid) that regulate many bodily functions. Often referred to as “the stress hormone”, aiding stress response is only one function it fulfills – it also plays a part in immune, metabolism, blood pressure, blood sugar and inflammation elements.

[20] This timing is key – cortisol levels tend to drop significantly (to about half) within about an hour of elevation – but if a dog is continually stressed (to the point of “distress”) then cortisol can continue to be released and take days to dissipate. If the dog is exposed to stressful events during that gradual decline, then its reactions can be “pre-set” to escalate.

[21] Hekman, J. P. et al. (2012) Salivary cortisol concentrations and behaviour in a population of healthy dogs hospitalized for elective procedures; Appl. Animal Behavior Sci; 141 (3-4); pp.149-157; research suggests future construction of a rating tool that identifies particular dog behaviours and links them to salivary cortisol concentrations – one aim being to develop effective methods to ameliorate stress in dogs.

[22] Marza, S. M., et al. (2024) Behavioural, Physiological and Pathological Approaches of Cortisol in Dogs; Animals 14; 3536; URL https://doi.org/10.3390/ani4233536 

[23] Schobert, I., et al. (2017) Psychobiological Factors Affecting Cortisol Variability in Human-Dog Dyads; National Library Medicine; Biotech; 18pp; URL https://pubmed.ncbi.nlm.nih.gov/28178272/  

[24] Petersson, M., et al. (2017) Oxytocin and Cortisol Levels in Dog Owners and Their Dogs Are Associated with Behavioural Patterns: An Exploratory Study; Frontiers in Psychology V. 8; Art. 1796; also determined that oxytocin and cortisol levels in the dog-guardian pairs studied are associated with the way the owners interact and the behaviours flowing from those interactions.

[25] Quote from Burman, O. (2014) Do Dogs Show an Optimistic or Pessimistic Attitude to Life?: A Review of Studies Using the ‘Cognitive Bias’ Paradigm to Assess Dog Welfare: The Social Dog - Behavior and Cognition; Ch 12; pp. 347-372. 

[26] AVSAB reference de Castro ACV, Barrett J, de Sousa L, Olsson IAS (2019). Carrots versus sticks: The relationship between training methods and dog-owner attachment. Appl. Anim. Behav. Sci.; 219:104831; URL https://www.sciencedirect.com/science/article/abs/pii/S0168159119300127 

[27] However, refer Mendl, M., et. al. (2009) Cognitive bias as an indicator of animal emotion and welfare: Emerging evidence and underlying mechanisms; Applied Animal Behaviour Science Volume 118; Issues 3–4; pp. 161-181. And see Richard Dawkins Foundation; URL https://richarddawkins.net/2018/08/the-cognitive-biases-tricking-your-brain/ 

[28] Parihar, V. K., et al. (2009) Predictable Chronic Mild Stress Improves Mood, Hippocampal Neurogenesis and Memory; Mol. Psychiatry; 16(2): pp. 171-183; available through Nat. Libr. Medicine Biotech Inf.; URL https://pmc.ncbi.nlm.nih.gov/articles/PMC2891880/: note also avail. at Mol Psychiatry - Author manuscript; in PMC: 2011 Aug 1; URL https://www.nature.com/articles/mp2009130 (hence date ref. of 2011 vs 2009).

[29] Recall that chronic means “persisting for a long time or constantly recurring”.

[30] Specifically new neuron growth in the hippocampus; also referred to as “hormesis”, viz., is a biological phenomenon where a low dose of a potentially harmful stressor, such as a toxin (drawing on the King Mithridates habit of ingesting poisons) or environmental factor, stimulates a beneficial adaptive response in an organism. In other words, small doses of stressors that would be damaging in larger amounts can actually enhance resilience, stimulate growth, or improve health at lower levels [this simple definition is from https://en.wikipedia.org/wiki/Hormesis].

[31] The Parihar, et al. (op. cit.) study infers distinguishing detriments/benefits from negative forms of stress (unpredictable) versus positive stress (predictable day-to-day stress) – question is, what does this tell us about using aversives in dog training?

[32] Refer Article on this website, “Reinforcement vs punishment: A dog trainer’s perspective” for a discussion of this point.

[33] Cain, C. K. and Le Doux, J. E. (2008) Chapter 3.1 Brain mechanisms of Pavlovian and instrumental aversive conditioning; Handbook of Behavioral Neuroscience; V 17, pp. 103-124; URL https://www.sciencedirect.com/science/article/abs/pii/S1569733907000070 ; Quote - Fear learning can be broadly divided into two categories: the acquisition of fear reactions, modelled by Pavlovian conditioning, and the acquisition of fear actions, modelled by instrumental avoidance/escape conditioning: and see Krypotos, A. M. et al. (2015). Avoidance learning: a review of theoretical models and recent developments. Front Behav. Neurosci.; 21;9: 189; URL https://pmc.ncbi.nlm.nih.gov/articles/PMC4508580/ ; Quote - Avoidance is a key characteristic of adaptive and maladaptive fear: Note that the definitions are – “in the context of human pathology, fear responses can be maladaptive if they are activated by innocuous stimuli or are activated in a recurring, excessive, or prolonged way that goes beyond the requirements of the situation”.

[34] Bekoff, M. (2017) Dogs Think About and Plan for the Future, Don't They?; URL https://www.psychologytoday.com/us/blog/animal-emotions/201712/dogs-think-about-and-plan-the-future-dont-they; Coren, S. (2017) Do Dogs Think About and Plan For the Future? URL https://www.psychologytoday.com/us/blog/canine-corner/201712/do-dogs-think-about-and-plan-the-future

[35] Thorndike, Skinner, Pavlov, Most and Koehler all referenced in an Article on this website Reinforcement vs punishment: A dog trainer’s perspective”.

[36] Note that the terminology of “stimulus discrimination” (or just the words “stimuli” or “discrimination”) is commonly used in the errorless learning context, because the early experiments in the field trialled pigeons pecking at a lighted button, where the correct/desired button was made brighter than the wrong/undesired button [Terrace, H. (1963a). Discrimination learning with and without “errors”. Journal of the Experimental Analysis of Behaviour, 6(1), 1–27; Terrace, H. (1963b). Errorless transfer of a discrimination across two continua. Journal of the Experimental Analysis of Behaviour, 6(2), 223–232].

[37] Handley, K. et al. (2023) Comparing trial-and-error to errorless learning procedures in training pet dogs a visual discrimination; Learning and Motivation Journal 84, 101944.

[38] Sidman, M (2015) Errorless Learning and Programmed instruction: The Myth of the Learning Curve; European J Behav. Anal. V. 11; Iss. 2. URL https://doi.org/10.1080/15021149.2010.11434341 

[39] Sidman, M. (1989; Rev 2000/2001) Coercion and its fallout; Publ. Boston, Mass.: Authors Cooperative; 300pp

[40] Avery, C.; Martin Seligman’s Contributions to Positive Psychology: A Comprehensive Overview; summarised at URL  https://achology.com/psychology/martin-seligman-and-his-contributions-to-positive-psychology/ 

[41] “I just wanted my pet dog to understand some basic manners – I didn’t want to learn how to punish it”.