Updated: Mar 15, 2020
As a species, humans tend to be judgmental of behaviour, often labeling behaviour on a continuum of good to bad. We have notions of right and wrong, morals, principals and social norms that define appropriate behaviours, all of which are, of course, important to ensure social cohesion and functioning of a civilized society. This means we are naturally inclined to have a bias towards judging and defining behaviour, although it is important to recognize those judgments often have more to do with our own narratives than what is actually happening. This naturally promotes an approach of either dealing with unwanted behaviours in a punitive manner or, using more positive means, encouraging a different more ‘appropriate’ behaviour in place of the other. In other words, we can tend to have an operant bias, which leads us just to focus on changing the behaviour without consideration to what might be driving it. There are important questions here though; What was the original behaviour communicating (actively or passively) in the first place? What emotional load might have been triggering the behaviour?
In human psychology there has been a movement away from purely working on the consequences of behaviour, and more of a focus on the emotional drive that comes before it. It is recognized that behaviour is emotionally driven, and consequence led. In the dog world there is still a big focus on the consequences with little regard for anything else.
If we are training a dog to teach new behaviours, then a purely operant model might be understandable. We must recognize, though, that even when training we risk looking to manipulate behaviours with little attention to the animal’s emotional state whilst doing so. For example, we can easily risk using the dog’s emotional drive to food to get a behaviour we want at the expense of the dog’s needs or wants. It can be easy to apply pressure through food, attention, play, etc.
It is important to recognize that punishment (or indeed reinforcement) can only be relevant to the individual if processed as such internally. For example, if someone is sent to prison, that punishment can only really be truly effective if the person emotionally processes that experience and has remorse for their previous behaviour. Or, the punishment needs to be so extreme that fear of the punishment is greater than the need/drive to give the behaviour, but this is always open to a return to the behaviour if that fear subsides or the emotional need increases. We know the inadequacies of just having a punishment based judicial system without addressing the factors that contributed to the offending. The same considerations apply to punishing animals.
If we are looking at supporting a dog with behavioural issues, we need to pay less attention to what they are doing, and more on why they need to do it, especially in regard to the all-important need for relief.
When looking at behaviour through purely an operant filter we run the risk of not recognizing the emotional load that triggered the original behaviour, or the relief the animal is seeking by doing it. The concept of relief is one of the most important concepts to consider when we work with dogs with challenging behavioural presentations. Stress related responses are more likely to be reflexive, less rational- more emotional. We can manipulate behaviour all we like, but if we are not providing relief for the initial emotional load, we risk building and supporting behaviours that we might find more appropriate but of little help to the animal. This is especially important when we recognize the extra burden put on an animal that is in pain or suffering from past trauma. (I will discuss relief in more detail in a further post as it is such an important consideration).
For example. Something as simple as a dog that jumps all over guests that come to the home. We might judge jumping as ‘bad behaviour’ however it is just a representation of the dogs emotional over arousal. This may well be excitement, but it can also be about social conflict and anxiety. If we work hard just to reinforce a ‘sit’, for example, as an incompatible behaviour, the unwanted jumping might be reduced or stopped, but what has happened with the initial emotional load experienced by the dog? Are they getting relief for that emotional load by being asked to sit? Are we creating extra internal conflict in the dog by reinforcing one behaviour but not addressing the initial emotional needs? What about the dog’s need to find ways of gathering social information in a ‘safe’ way, often hidden in a dog’s initial enthusiastic greeting? Also, often the dog sitting is seen as the dog being ‘calm’- but is it? Does the dog functionally putting their bottom on the floor and staying there truly denote a nervous system at rest? Often, wrongly, owners and professionals can class a dog as ‘OK’ just because they are not doing anything. If we are classing a dog as ‘OK’ purely based on lack of observable behaviours, then fair enough, but to make a judgement from that as to the dog’s emotional state is an error and sometimes a dangerous one. There can be many considerations behind something we feel some simple training can deal with.
The definitive statement that is often given is ‘behaviours that are reinforced will be repeated'.The power of this leads to a fine line between manipulation and support. If that behaviour given naturally by the animal offers the emotional relief, then great. If, through training and support, we can gain a more ‘appropriate’ behaviour that can deliver the same relief – excellent. If we are manipulating that behaviour purely because we like a new behaviour more, or we find it more appropriate, then we need to think carefully.
Emotions will always find a way out
There is a great saying in human psychology – ‘emotions will always find a way out’. What this means is that unless we get relief for our emotional responses, that response may find a way of coming out in other ways. In the example above, if the jumping up was actually a manifestation of an appeasement response to the social pressure and anxiety they are experiencing, getting the dog to sit may not help that emotional load at all. I recently worked with a dog where the dog’s greeting behaviours towards guests had been dealt with in a very operant way (sit and stay) but now the dog really struggled to settle with guests arriving and has started barking more during the visit. The dog had been reinforced to do something ‘physical’, but no work had been done to help the dog’s emotional state around guests to the home, or to understand the dog’s individual needs for safe social processing of visitors. Often ‘reactive’ behaviours are a consequence of the nervous system engaging ahead of any real cognitive processing of the environment. Once that nervous system engages then relief is needed and relief seeking behaviours begin.
If we were able to pause a behaviour and go back in time a little, so much has happened neurologically and physiologically- this is sensory integration. The brain has gathered information from its external senses, and importantly its internal sensory system (pain receptors, nervous system etc). It then cross referenced that with information already in the brain, pushed through various filters and then acts on it. When we are sufficiently calm enough, we can moderate responses with rational input, but if the nervous system activates too strongly then there is limited ability to influence what will more likely be a reflexive response. So, the final behaviour carries a lot of ‘load’ before it is physically manifested. There is a real risk this diverse and complex series of internal processes can be too readily discarded with a purely operant viewpoint. (For lots of geeky canine neuroscience check out the brilliant Kathy Murhpy's FB page Barking Brains).
We have some wonderful people in the industry, such as Sarah Fisher at Animal Centred Education, who are encouraging professionals and owners to look more in depth at what is happening pre-behaviour. Doing so helps us observe behaviour in a less judgemental (right/wrong or appropriate/inappropriate) and more ‘holistic’ way. We all need to allow ourselves more time to investigate what the dog might be experiencing emotionally and physically, what that behaviour is communicating to us and, most importantly what ‘relief’ the animal is trying to achieve. Observing body language and slowing down the environment to get more data along with the use of free work (check out ACE Connections on Facebook) can be really helpful in doing this. Equally important is learning more about the dogs preference regarding social and environmental processing (I will be writing a future blog on this topic). Individual preferences regarding processing form an important part of sensory integration. We should stay mindful that sometimes asking the dog to respond to lots of operant training and cues can hamper this ability to naturally process the information they need from the environment and/or orientate to feel comfortable.
Pitfalls of an operant model
A side effect of the over emphasis on operant models is that it leaves room for the constant discussion and debates regarding methodologies. In an operant model, all methodologies have equal legitimacy. Operant thinking puts the emphasis on the observable increase or decrease of behaviour, and this creates a biased view on how we perceive results. Even promoting more positive methods still puts the focus on the behaviour being the problem. This pursuit of results based on behavioural output alone can easily lead to a creep to more punitive methods, as supported by a LIMA model (least intrusive, minimally aversive). For example, look at a case of a dog reacting to another dog. If we simply define results as reducing the presenting behaviour, then punishing can be equally as effective as reinforcing another behaviour. However, punishing does nothing positive to address the emotional load that triggered the behaviour in the first place or the relief being sought by the dog. However, if we start to measure results in terms of the relief of the animal for that emotional response, punishment cannot be seen as effective. It is worth noting that positive reinforcement is equally capable of suppressing natural behavioural output; we must always stay mindful that we may be getting a behaviour we find more appropriate but doesn’t offer the relief the dog needs!
The Last Resort
Often, when positive methods seem to not get the ‘results’ required, the old argument of using more punitive measures come in to 'save the dog' as a last resort. The narrative here is that in order to satisfy the focus of reducing the unwanted observable behaviour, harsh, painful consequences are sometimes justifiable if positive methods have not. I pose this question - Would we have the same mindset when working with children? Would we expose a child with anxiety to the trigger they struggle with and if we can't help them positively then consider punishing their stress response instead? Or, if we can't identify or resolve the causation of the child's apparent poor behaviour would we resort to corporal punishment instead? The answer is, of course, no.
In relation to dogs, better management is always the best last resort and often the first choice, whilst a process is gradually and thoughtfully followed with empathy and compassion for the animal.
This is the big risk with an operant bias – it can shift us to associating results purely with what an animal does with little consideration for how they are feeling. I understand that some owners/carers can put pressure on professionals to get quick results, but that shows the importance of managing clients expectations and helping to adapt and change client narratives, more than it is an excuse to use harsher methods. Helping people to understand what their dog is experiencing is the best way to support their role as caregivers.
The paradox of the positive trainer
There is another angle to the operant bias, and that is the behaviour of those within the ‘positive’ side of the industry. It is sometimes noted that some people who practice positive reinforcement can also be judgmental of others, posting hurtful comments that can sometimes be seen as bullying etc. Commentators find this a paradox – how can those who practice positive reinforcement when training be so difficult and horrible to other trainers? Yet this is because we are seeing it in operant terms again. If people lack empathy, compassion or emotional maturity then regardless of what training method they use they will be inclined to post negative and sometimes hurtful comments. Using and preferring a certain method from the operant choices available indicates nothing, sadly, of a person’s emotional considerations of others or capacity for emotional empathy.
Part of the problem is that for too long we have seen animals as being subservient to us, with their behaviour being defined, managed and manipulated through the filter of a human's needs. While positive reinforcement has offered us a kinder, more humane way to train animals we should be mindful of the adage ‘just because we can doesn’t mean we should’. More importantly - just because it works, doesn’t mean it’s good. We must always stay mindful of the power we hold to manipulate contingencies and influence consequences.
This piece is not meant to be ’anti-training’, far from it, however as we learn more about the neurology and physiology of the dogs in our care it is our responsibility to insure we pay attention to their emotional needs and the relief they crave through their natural behaviours.
Educating the public
Educating the public about the amazing, emotional, sentient, feeling creature in their care is just as, if not more, important than talking about the right and wrongs of certain tools and methods. Once an owner gets their dogs emotional needs and the relief they seek, they will naturally be less inclined to use punitive means. If we focus on the operant we risk viewing the behaviour as the issue – the thing that needs to be changed regardless of what methods are employed to do so. There is also the risk of applied pressure to the carer with constant talk of 'train more' - for some dogs that is the last thing they need and can make the carer feel they are to blame as despite how much they 'train' the dog makes little or no improvement.
If we can put the focus on educating owners that the behaviour is communicating emotional need rather than something in itself to be changed, owners may be more inclined to listen to it more. Most importantly, if they come to see the behaviours as relief seeking, they are more likely to want to help offer that relief.. They may even recognize that sometimes allowing the avoidance of triggers is not a ‘cop out’ but the signs of a great carer.
This is not just about recognizing the increasing body of evidence to support the emotional drive to an animals behaviour, but equally about encouraging more empathy and compassion into the centre of everything with do. A less judgmental view of behaviour will allow a truly dog centered care approach that encourages us all to listen to the communication offered through behaviour and become less driven to simply want to change it.
Andrew Hale 2019
Train Positive Dog Behaviour & Training