This chapter covers
Definitions of obedience
Difference between training and behaviour modification
Reward versus punishment in canine versus human perception
Well-behavedness versus obedience
The client and I stood in the consulting room – she trying unsuccessfully to restrain her excited Springer Spaniel and me, equally unsuccessfully, attempting to examine him. This was in the days before I understood that such behaviour on the part of the patient was the ‘faff-about’ response to threat and stress, whereby he was seeking to repel unwanted attention by wriggling vigorously. Unbeknownst to me at the time, had I and his owner persisted in forceful and coercive handling, the dog may well have been left with no other option than to snap and bite.
As it was, I simply labelled him a ‘badly-behaved’ dog who was wasting my time and stretching patience to the limit.
The owner, sensing my tacit disapproval, retaliated, ‘He’s a highly clicker trained show dog, you know!’
By this stage in my behavioural career, I was aware of what ‘the clicker’ was and how it was supposed to function, but I resisted the temptation to ask why it had been left at home if this had been the primary means whereby the dog was accustomed to being given information. But it was the revelation that he was a show dog which proved most helpful. I knew that, having attended ring craft classes, a solid ‘stand-stay’ was a prerequisite for the show ring and being examined by judges. By the same token, I was told that a ‘sit’ was a complete ‘no-no’, as the last thing a show dog should do was park his bottom the minute a judge’s hands got anywhere near the rump.
The ‘stand’ command was duly given by the owner and understood by the dog to mean ‘Let this stranger run their hands over you without moving’. Lo and behold, the dog obeyed. From then on, the examination proceeded smoothly and I was able to commend the client on how well the dog was trained. However, one could almost imagine the dog himself raising eyes to Heaven at the stupidity of mankind and asking, ‘Why on earth didn’t you just say what you wanted me to do in the first place?’
Illustration by Victor Ambrus.
Many valuable lessons were learned during this potentially disastrous encounter which really focused my mind upon the true nature of obedience and how easily we can confuse dogs by our actions and clumsy communication. Dogs after all are responding to, or ignoring, environmental events all the time in whichever way they see fit. Why do we imagine that what we say and do are so special? And if a dog appears not to listen or understand, whose fault is it?
What is obedience?
Standard definitions of the word ‘obedience’ are all along the lines of the following, readily provided by Google dictionary:
compliance with an order, request, or law, or submission to another’s authority
Others include the following:
•‘Willingness to obey orders’ Chambers Pocket Dictionary
•‘Submission to another’s will’ Oxford Dictionary and Usage
•‘Complying with rules’ Penguin English Dictionary
•‘Doing what you are told’ Oxford School Pocket Dictionary
It is hardly surprising therefore that when asked to say what obedience is and to apply it to dogs, the average owner universally replies with utter confidence and no small a hint of ‘what a stupid question!’: ‘It means he must do as he’s told!’ The implications of this, at best, oversimplification and, at worst, damaging misunderstanding are several and profound. They are the cause of much conflict and mismatch between human and canine expectations.
Dogs are first and foremost expected to know that there are rules and what those rules are. Obedience is equated with submission and they are expected to defer to an authority. They are expected to know what words mean. Obeying ‘orders’ implies the involvement of threat and coercion in carrying out the order. All these interpretations feed readily into the concept of dominance and hierarchy. Whether we are conscious of the fact or not, implicit in all the definitions is that an obedient individual would really rather be doing something else. In order to be obedient, therefore, they must lose out.
This inevitably brings dogs and people into conflict. For the health of the dog-human relationship in general, and when counselling individuals, it is always essential to thoroughly clarify exactly what we all mean by ‘obedience’ before attempting to tackle the problem at hand. It may even be that, once this is explained, owners can begin to answer their questions and to solve ‘problems’ for themselves.
My own definition is this:
An obedient dog is one whose own choice of behaviour can be successfully altered to coincide with that of the owner.
This definition emphasises both that the dog has a choice in the matter and that the owner must have determined in advance exactly what they want the dog to do in order to instil obedience. It does not, however, indicate how the dog’s mind should be changed.
It is worth pointing out at this juncture that the owner is not the only source of information to which a dog may respond. Yet all too often, owners assume that they are all-important and that, in every circumstance, they should be the most important item on the dog’s horizon. Broadly speaking, obedience training refers to information and instruction emanating from the owner only, as usually happens in the standard training class. Behaviour modification, on the other hand, although inevitably involving owner-centred training, acknowledges the effect that other features of the environment have on a dog. Such influences may inadvertently ‘train’ a dog as well, if not more efficiently, than any process applied deliberately. In behaviour modification, environmental events as well as owner behaviour are manipulated to effect behaviour change in the dog for the better.
Let’s take a typical example of ‘environmental training’: the dog that continues to bark at passers-by despite their owner’s efforts to stop him. A dog is not to know that ‘passers-by’ are going to pass by anyway without his intervention. Inevitably, when they do, his barking is rewarded by the potential intruder’s retreat and its apparent success. Dogs may deliberately seek out and remain in the best ‘look-out’ position (often the back of the settee next to the window) in order to be ‘successful’ as often as possible.
On what basis do dogs choose behaviours?
We are all familiar with the ‘carrots and sticks’ metaphor with respect to guiding behaviour, whether of donkey, dog or child, but are perhaps not as au fait as we should be with what forms each can take and why they ‘work’. According to the definition above, the outcome of the application of carrot or stick ought to be the exact behaviour we want the dog to perform. This is straightaway at odds with what the average dog owner needs, which is simply to find a means of stopping the annoying stuff. The human default position with regards to altering any undesirable behaviour of fellow man or beast is always to try to prevent it, generally by some form of punishment. Behave or else! Sticks may appear to have instantly gratifying results, but do they teach the dog what to do instead of being annoying? If used indiscriminately, without forethought or understanding, they may seriously backfire.
What are the means whereby outcomes for a dog are created and, with a modicum of luck, learned? Reward for getting it right or punishment for getting it wrong? It comes as a sobering thought when pointed out that the dog training fraternity have over time been far more inventive when it comes to sticks than carrots. Choke chains, correction sprays (human and bark activated), electric shock collars and so-called ‘Freedom Fences’ (buried electric shock fences) are all intended to punish wrongdoing while the humble food reward for getting it right needed no lucrative reinvention.
It must also be taken into account what outcomes have a rewarding or punishing effect as far as the dog is concerned. We know what would be nice or nasty from our own perspective, but not necessarily from the dog’s. A hug from the owner may be enjoyed, but not if a child or stranger tries to do the same thing. An owner’s disapproval may be highly emotionally significant whereas that of a stranger simply viewed as a threat. The value and impact of rewards and punishments also vary according to context. How much does a dog want to gain a reward or avoid a punishment at any particular moment in time? In the consulting room, close contact with the owner may override food as a motivator and admonishments pale into significance compared to the desire to escape a vet’s clutches.
A Little bit of theory
Reward and punishment
Reward is defined as ‘the consequence of an action which renders that action more likely to be repeated’.
Punishment, on the other hand, is defined as ‘the consequence of an action which renders that action less likely to be repeated’.
In addition, both rewards and punishments can be positive and negative. In positive reward, something desirable is gained. In negative punishment, something desirable is lost. In positive punishment, something unpleasant is experienced. In negative reward, the unpleasant stuff stops.
Clear as mud? Confused? This is not surprising especially when the sort of punishments we consider to fit the crime for fellow humans may be, technically speaking, nothing of the sort. A punishment can only be defined by result, that of a reduction in crime, not by how many weeks one spends behind bars or how big the fine is. Both these are negative punishments, as the perpetrator is deprived of freedom or money in order to teach him a lesson. Retributive, ‘an eye for an eye’ or ‘tit-for-tat’ behaviour is not punishment.
The nature of a punishment or reward can only be evaluated by the dog itself and its success or failure judged by the behavioural result.
It is frequently necessary to clarify this to owners – that it is only how a dog sees things that matters, not what we think should be significant. It may be assumed that punishment only occurs by physical means, by smacking for example, and that ‘just shouting’ cannot constitute proper punishment. Yet there is something that may have a far more potent effect on our canine companions than is generally realised: human attention. We must not underestimate the importance of how it is given or withdrawn, when it is given or withdrawn and for what reasons, and the consequent behavioural impact.
Dogs enjoy many things other than food which, in the right context, can be used to ‘train’ them. The supply or withholding of pleasant as well as unpleasant experiences can be used judiciously to strengthen behaviours we want and weaken those we don’t. Of the essence is the word ‘judicious’.
The problem is the confusion in people’s minds, including those of trainers, as to what is really happening in the mind of the dog. It is easy for most people to understand that positive reward, for example in the form of a piece of hotdog, given to a dog for sitting when asked, will act as a reinforcer and increase the likelihood and frequency of sitting occurring. They may also grasp that they must withhold the said hotdog until the dog sits. But do they realise, that in doing so, they are negatively punishing all other behaviours, other than the sit, that the dog tries in the meantime? And that the withholding of a specific intended reward will only ‘work’ if accidental reward for the wrong thing is not coming from elsewhere in the meantime?
Even less well understood is the relationship between punishment and negative reward (or reinforcement). Is it more important for a positive punishment to stop a behaviour (sheep chasing for example) or for the behaviour which appears to cause the punishment to stop to be negatively reinforced? There is lack of appreciation of the role of negative reward in our routine daily interactions with dogs as well as when purposefully dealing with the extreme of sheep chasing. Yet if these principles are fully understood, it is easy to see how negative reward can be created in the mind of a dog out of almost every unpleasant experience. This is achieved, firstly, by being fully aware of what the dog will find unpleasant and, secondly, by deciding which behaviour the dog must perform to have the effect of stopping the unpleasantness, thus bringing the potent emotional reward of relief.
Great skill and understanding is required regarding how each pressure is brought to bear upon behaviour in order to use them deliberately and why manufactured punishing devices (including those utilising electric shock, ultrasonic sound and compressed gas) can be so easily abused. The accompanying graph illustrates the difference in use of the two pressures (punishment and negative reinforcement) using electric shock as an example.
Do not for one moment imagine that punishing devices are being advocated here. I can think of nothing worse than deliberately putting a dog in a context in which they are sure to misbehave and then using some potentially painful gadget around their neck to punish them when they do so. The very idea is abhorrent. But understanding these principles can be used to explain why lesser aversives, such as getting cross or smacking, which happen on a daily basis, may seem to be effective at the time. Yet if getting cross ‘worked’, why do dogs keep repeating behaviours which make us cross? Why do they not learn what they should do instead? This is what mystifies owners.
Dog A belongs to the average dog owner. All that the owner requires is for a particular behaviour (or several behaviours) to stop and stop as soon as possible. The dog is typically put in surroundings in which the behaviour occurs at its worst. The punishment is then applied at an intensity sufficient to interrupt the behaviour, at which point the punishment ceases. No thought is given to what the subsequent behaviour should be.
Dog B is in the hands of a skilled user of punishing devices. Rather than simply stopping a behaviour, a desired target behaviour is aimed for. The dog is put in the least arousing situation to begin with and all behaviours other than the target behaviour are punished. As soon as the target behaviour is performed, the punishment ceases. It is therefore rewarded by absence of the aversive.
For both dogs A and B, the motivation for the unwanted behaviour (or how much the dog is ‘driven’ to do it) must be taken into account. Dog A is typically put in a situation where the motivation is very high (the dog is already chasing sheep for example). If the dog is just looking in their direction but not chasing them, there’s nothing to ‘punish’ yet, right? Wrong! By the time chasing has begun, the dog will be so aroused that the intensity of punishment required to interrupt chasing will be correspondingly very high. There is therefore a great risk of unacceptable very painful levels of aversive being proportionally applied. Even so, the dog may override pain in favour of chasing, resulting in multiple applications of shock.
Dog B may be just looking at sheep or walking slowly towards them. The intensity of punishment required to ‘change the dog’s mind’ will be correspondingly low but will continue until the target behaviour is reached (turning to look at trainer for example). The risk of failure and of inadvertently causing pain or distress is far lower.
Another way to explain the emotional effect of reward and punishment is to use the accompanying Scale of Emotion diagram. This diagram describes (in much simplified form) that all training methods work by producing positive emotional change, whether in the form of relief at the avoidance of punishment (negative reward) or satisfaction in gaining a resource (positive reward). It also emphasises that all training is more effective if a dog is in emotional ‘neutral’ to begin with.
However much one disapproves of deliberate punishment in practise, we cannot deny the theory underpinning its proper use and that, if used knowledgeably with excellent timing, it can be effective. We also cannot deny that the human default value is to get angry, raise voices, threaten and smack and that being dealt with in this way is what domestication has led the dog to expect. Even so, we should not condone the deliberate application of potentially abusive techniques and devices. Rather we must understand the principles that govern their use, including the power of negative reinforcement, in order to explain to clients why the things they do instinctively when dogs misbehave may appear to work or may go disastrously wrong.
For all undesirable behaviours, there are three ‘golden rules’.
1.We must have a clear idea of what we want the dog to do instead and thoroughly train that behaviour.
2.We must not put our dogs into such difficult situations that they misbehave and we feel forced to shout louder, jerk the lead harder or ‘up the ante’ on an electric shock collar remote control.
3.We must positively reward a dog when he gets it right, however long it takes and however angry they have made us feel.