“If dressage is the highest expression of horse training, dressage judging should be a transparent and systematic system for de-constructing that training”. Horses and People Editor, Cristina Wilkins, speaks with Andrew McLean about his proposal to align the dressage judging system with evidence-based horse training (learning theory) as a means to making results more objective and transparent.
Throughout this exclusive series, Dr Andrew and Manuela McLean, founders of the Australian Equine Behaviour Centre (AEBC) and Directors of Equitation Science International, will explain dressage judging against the backdrop of learning theory and the equitation science training scale. Just like step-by-step horse training, breaking down the dresssage judging criteria into chunks (training deficits) will help you identify and rectify specific training deficits, so you can design clear training strategies to enhance your next competition performance.
Dressage is training, and training is learning
Dressage derives from the French word ‘dresser’ (meaning ‘training’) and, in its ultimate form, is considered the highest expression of horse training. At a national level, official competitions are governed by each country’s equestrian federation (e.g. Equestrian Australia). Their rules are based on those set by the International Equestrian Federation (FEI).
The FEI rules state that “the object of Dressage is the development of the horse through harmonious education,” and should ultimately result in “achieving perfect understanding with the rider.” There is a clear premise that a horse’s education or training progresses in a series of ‘steps’ and this is reflected by each of the different levels of competition, from preliminary to Grand Prix, whereby horses have to demonstrate they can perform at one level before moving on to the next.
The FEI concepts of ‘harmonious education’ and ‘perfect understanding’ imply that the horse has learned how to behave (respond) to the rider’s aids, whereas ‘development’ refers to the physical changes to the horse’s body that allows the horse to carry a rider in balance with increased strength and power. This ‘education’, therefore, has two components; one is mental/cognitive (the psychological process of learning new skills) and the other physical (the gymnastic development and strengthening of the muscles).
In the past, most of the schools and systems of riding have focused mostly on how the rider communicates with the horse; that is, if you ride well enough, your horse will perform well. However, by focusing on the rider’s ability to deliver the signals/aids, they do not promote a clear understanding of how the horse learns to respond correctly and fluently to the rider’s aids in the first place.
A good seat and accurate application of the aids are obviously important – they allow the rider to be very clear and consistent – but the horse first has to mentally absorb, process and retain the meaning of each of the rider’s cues (the correct response for each aid).
The cognitive process of learning has been studied extensively in behavioural psychology and is known as learning theory.
How the horse learns
Dr Andrew McLean was the first to describe how to explain dressage training using learning theory. Working in collaboration with Prof Paul McGreevy , a specialist in veterinary behavioural medicine and animal welfare science who has assisted Andrew to disseminate this knowledge to the scientific world, the pair have refined many aspects of learning theory in horse training. Together with his wife Manuela, Andrew has developed a complete and practical training system that aligns with learning theory and respects the horse’s mental abilities, instinctive tendencies and biomechanical potential. This system is described in detail in the book ‘Academic Horse Training’.
In great demand as a trainer, coach and speaker, Andrew travels the world explaining to both lay and academic audiences how training works at all levels in a refreshingly simple, logical and illuminating way.
Shaping scale in learning theory
Andrew and Manuela developed the shaping scale of horse training (To view the shaping scale history review and explanatory tables we recommend you download the pdf version of this article). Known as the Equitation Science Training Scale, it is a systematic progression that identifies six stages of learning to perfect each response, from a basic attempt (an approximation of the correct response) to proof (where the horse immediately responds correctly every time and in any environment).
The scale is based on the principle of shaping: reinforcing successive improvements that are approximations of the final response. Effective trainers have always known how to ‘shape’ behaviours – they analyse the components of the behaviour, and break them down into small chunks the horse can understand and store. Like steps on a ladder, each chunk leads on to the next and, over time, the behaviour is perfected.
By rewarding the horse often at the right time, great trainers capitalise on each tiny improvement and move quickly towards the final response.
In fact, Dressage Master Gustav Steinbrecht says in his book ‘Gymnasium of the Horse’: “[The training exercises] should all follow one another in such a way that the preceding exercise always constitutes a secure basis for the next one. Violations of this rule will always exert payment later on…”
Dressage training versus judging
Recently, Andrew has turned his attention from training to dressage judging. Speaking at several 2014 international forums, including the International Equitation Science Conference and the Global Dressage Forum, Andrew put forward a proposal to make dressage judging more objective and clear by applying learning theory.
“If dressage is about training,” says Andrew, “and training animals is about using learning theory to shape the criteria of responses, then dressage judging is about de-constructing training; de-constructing what the trainers are trying to do.”
Judging in any sport should be about objectivity – facts that are probably true – rather than subjectivity, which are opinions, personal interpretations, unsubstantiated assumptions and inherent biases – but many people believe dressage is rife with subjectivity and prone to inherent biases.
Dressage professionals have highlighted problems, such as discrepancies of 30-40% between individual judges and inherent bias. In an article published by eurodressage.com, CEO of the International Riders Club and FEI competitor Wayne Channon called for judging to move into the 21st Century and nominated four levels of judging bias:
- Conformity Bias – judges feeling that they have to fit in with their colleagues.
- National Bias – scoring one’s own nation higher than the other judges.
- Order Bias – the later a rider starts the better chance they have of a higher mark.
- Memory Bias – remembering the previous marks awarded to the same horse/rider combination and awarding marks based on that performance, rather than what was actually seen on the day.
Channon points out that judging needs to be simple so judges get it; transparent so riders, coaches and the audience get it; objective so riders can trust it and; reliable so that what we value is not too complex.
Importantly, Channon declares himself “a proponent of separating the judging tasks into manageable chunks” – a statement which Andrew points out sounds just like shaping.
Far from attacking the judges, Andrew believes judges judge what they see at the time, based on the FEI training scale. His review proposal centres around enhancing the FEI training scale by making it truly scalar and aligning the elements with learning theory. In this way, the scale could be applied to every dressage movement and used as the marking criteria.
Establishing training priorities
Equitation science emphasises that horse training should prioritise, for the sake of safety, obtaining clear control of the horse’s mobility, which should be achieved from light aids to protect the horse from strong or constant pressures. This, as Andrew reminds, is old classical training and ideology – legs first, then frame; in other words, the head and neck posture is a consequence of training the legs to respond to our aids. Problematic head carriages almost always reflect major training holes in lightness to either leg or rein aids or both.
“All dressage movements are based on just four mobility responses,” says Andrew. “These basic responses, which have to be trained in all gaits, are the basic responses that horses learn and are reflected in the lower level dressage tests.”
Scroll down to download the pdf version of this article and see all the text boxes and images that were shared in the original print article.
Andrew explains that, in training, the rider first achieves a basic attempt of these four responses. By applying learning theory correctly, the horse then learns to respond immediately and from a light aid every time. This is what behaviourists call stimulus control, which in equitation may be better understood as ‘obedience’ – an immediate response to a single light aid.
Once these basic responses are installed and once the horse develops the physical strength required, the responses can be chained together one after the other (not at the same time preferably as it could confuse the horse) in cascades of responses to produce the more advanced and complex movements.
For example, half pass is a composite movement consisting of a turn response and a yielding the hindlegs response. Shoulder-in is a turn of the forelegs off the line, while travers is an indirect turn on the line followed by a yielding the hindlegs response. The quality of these higher composite responses depends on the quality of the basic responses.
Andrew insists that the first steps of basic attempt and stimulus control need to be added to the FEI training scale before rhythm.
“There is no slot in the FEI scale for basic attempt and stimulus control,” says Andrew. “It is assumed, probably because in the early days when dressage was first introduced as a sport, the horses, which were mostly military mounts, were trained and had already achieved rhythm level for the required movements. We can’t assume that anymore today. Horses may perform at competitions when they are only beginning to learn shoulder-in or piaffe and we need to be able to assess this and mark them accordingly.”
Stimulus control, which Andrew terms ‘obedience’, should be assessed as happening immediately from a single light aid. It is easy to see if the rider is constantly spurring or holding the horse, or if the response is delayed, and this ties in with the element of ‘rhythm’.
Andrew insists it is critical to define rhythm according to the oldest and most useful definition – that the horse keeps doing what was asked. Judges often differ in their interpretation of the term and some believe it is the regularity of the gaits. “While regularity is an important aspect of rhythm,” says Andrew, “it is secondary from the training perspective of continuing to respond to the single aid.” When seen from this perspective, rhythm can be directly observable and assessed with tests of self-carriage (momentarily softening the rein and/or leg contact). Self-maintenance of the response is what training is all about and, in Andrew’s view, it is essential to ensure good welfare.
Straightness is shaped after rhythm and can also be assessed with self-carriage tests. “Straightness,” says Andrew, “is not about the straightness of the body, but rather a function of laterality, and to a smaller extent, asymmetry. The effect of laterality is that one diagonal pair of legs tends to provide more propulsive thrust than the other. The other diagonal pair tends to provide slightly more reverse thrust and the difference between them skews the horse.” Biomechanically, ‘crookednes’ is normal for quadrupeds as it allows them to place the hindlegs deeper so that one leg goes in between the forelegs, as happens in jumping or galloping. “When training lightness and rhythm,” he continues, “if the trainer focusses on making the diagonal pairs as even as possible, the horse will maintain perfect rhythm and straightness.”
The ‘contact’ or connection level in the scale includes other subtleties, not just the horse-rider contact, but the further physical development of the horse’s impulsion, engagement, ‘durchlaessigkeit’ (throughness) and collection. Rather than being actual learned responses, these elements occur as the horse becomes physically stronger.
‘Durchlaessigkeit’, a German word that is as hard to translate as it is to pronounce is known as ‘throughness’, but has been better described as ‘letting go-ness’ or ‘when the rein aids flow through to the hind legs’. This, says Andrew, is a very important part of training a horse: “it is not good enough when you pull on the reins and the horse just slows his front legs. His hindlegs must also dig-in to some extent and this occurs by lowering the hindquarters.”
“When seen from this perspective, the degree of throughness also corresponds with the degree of collection because, as the horse improves and develops physical strength, the hind foot prints begin to step closer to the front foot prints. So, impulsion, engagement, throughness and collection should be seen as on a continuum at the same scalar level but varying only for the training stage.”
The last step of the equitation science training scale is ‘proof’ – an element that Andrew says he ‘borrowed’ from American dog trainers. “It is important to show that what you have trained at home in your indoor arena will still work when you go elsewhere,” he states. “Horses are very contextualised in their learning (they make associations between the behaviour and the place where they are doing it), so it is normal for the wheels to fall off to some extent when you move to a different environment. It is important to gradually shape that as well – taking them to slightly more challenging places over time.”
Merging dressage training and judging for better transparency
If the equitation scale was universally accepted to describe the training process, the judge’s marks would reflect the actual training error. This way, trainers would have a better idea of where the training had gone wrong and what elements need to be re-trained, improved or refined for each response or movement. By emphasising the importance of lightness and self-carriage in all movements, a scale like this would also align good judging with good welfare.
For example, instead of the current criteria for a mark of 5 being ‘sufficient’, the benchmark would become ‘basic attempt’ – the required movement was visible and happened at the correct place. Marks below 4 would indicate a failure at basic attempt and the severity would reflect the gravity of the training problem. Marks from 5 to 10 would align with the training scale to show the level of development towards harmony between horse and rider.
The scale could also be used to set ‘firewalls’ or ceilings for marks, so that if any step is not present, the mark is frozen at the step below. According to Andrew, rhythm should be firewalled, so that no horse, regardless how expressive his movement, can be marked higher than a 6 if he is not showing self-carriage. Straightness is another level that should be firewalled, not just because crookedness thwarts the horse’s ongoing progress, but also because the lack of it may mean the horse is subject to ongoing pressures on one side or the other. This would mean that a horse that is not straight could not receive marks higher than 7.
Andrew is a firm believer that judging and welfare should be synonymous. “Judges are the housekeepers of welfare and the system should be transparent for all. If we want to see dressage go into the future, it must be sustainable, it must move away from its cultural shadow and become more objective. Importantly though, a more objective system would give judges more confidence and credibility.”
Scroll down to download the pdf version of this article to see how the Shaping Scale could be applied to judging specific movements such as, the halt, lengthened strides and piaffe.
While the dressage world is entrenched in its own history and traditions, and changes such as Andrew proposes will take time to be accepted, the truth is that, with a good understanding of learning theory and the shaping process, it is still possible to make some sense of the current judging system against the equitation science shaping scale in a way that can inform your training.
In the next part, which will be available soon, Manuela McLean will begin to list the most common comments that judges make when scoring dressage movements and explain how they fit in with a learning theory approach to horse training, and describe strategies for rectifying problems. Comments such as ‘not shown’, ‘resistance’, ‘not forward’, ‘needs to be more uphill’, etc., will be grouped and placed into the appropriate level of the equitation science training scale. Clear training strategies to improve each response will be broken down and explained.
To find out more about Equitation Science International and the work of Dr Andrew McLean and Manuela McLean, visit the ESI website.