If we can define 'Harmony', can we score it?

nat waran
Aug 4
9 min read

Professor Natalie Waran OBE

For as long as people have ridden horses, harmony has been the word used when everything goes right - when the horse appears willing, the rider looks still and everything looks easy.

Dressage is now going through its first full rules revision since 2022, and the first draft of the FEI proposals for 2027 has been published. This has resulted in a good deal of the discussion about where the word harmony belongs on the score sheet.

British Equestrian has asked for a clear definition of harmony, suggesting that it should be built into the score for each movement rather than sitting in one collective mark at the end. Other organisations would prefer a second collective mark, with one version separating the training of the horse from the performance of the rider, and another separating the training scale from the harmony between horse and rider. The FEI dressage technical committee has gone in the opposite direction and proposed removing collective marks altogether, so that harmony and the training scale are judged inside every technical movement.

So, what should be done? Judging dressage is difficult enough because it relies on human judgement, and humans can be influenced by unconscious bias, by differences in interpretation, by the reputation of the rider and by what each judge has come to expect. Even where judges are highly trained, variation in how subjective criteria get applied can reduce transparency and affect public confidence. Other sports that are scored this way have had similar challenges, and the experience of gymnastics, ice dance and freestyle skiing suggests that a judging criterion only holds up when the definition is robust, the judges are calibrated against one another, objective measures sit behind the scoring and the whole system is open to scrutiny.

The current debate related to the rules revision raises important questions for equestrian sport. Can harmony be defined clearly enough so that different people watching the same test reach much the same conclusion, and can it be measured well enough for the mark to be both meaningful and defensible?

Is harmony art or science?

Harmony came into equestrianism as an artistic concept, and used that way it describes how a performance looks, with movement that flows rather than starting and stopping, signals between horse and rider that are hard to spot, and nothing that appears forced. Almost anyone who watches horses recognises the quality being described, which is why the concept has survived so long without more scrutiny.

Science asks a different question that's more about function than form – ie what produces that appearance, what causes it to occur? The rider gives signals that are consistent in shape/placement and timing and then takes the pressure off the moment the horse answers, while the horse, having learned what those signals predict, responds promptly and off an aid applied lightly so there's no force.

Neither description is wrong, but only the second one can be authenticated, while the first relies on the eye of whoever is watching.

Good judging should recognise not only what the horse does, but how it was achieved.

Breaking harmony into it's different parts

The way forward might be to stop treating harmony as one overall impression and start treating it as something built from components we can name/define/describe consistently and measure. It’s better understood as the appearance that occurs when several separate things line up at the same time rather than as a single quality the horse either has or hasn’t got. This is why one mark at the end of a test will struggle to capture the quality. It is suggested that Harmony be broken into four parts, or domains, each of which can be described, taught and observed (Figure 1).

1. Communication

The rider gives signals that are clear, consistent, well timed and suited to the stage of training the horse has reached, and the horse shows it has understood by answering promptly and predictably. Learning theory explains how this develops, and it gives us principles we can actually measure, particularly around the timing of the aids and the release of pressure.

2. Behaviour and affective state

Affective state is the scientific term for how the horse is feeling, and behaviour is our evidence for it. Conflict behaviours, facial expression, where the horse is looking, body posture and how quickly it settles after being asked to do something difficult all tell us something about that inner state. It isn’t only about observing negative indicators, but also looking for confidence, engagement, willingness and appropriate responsiveness.

3. Biomechanical function

This part is about how the horse moves and what its body can comfortably do, which comes from the interaction between the physical capability of the horse, the influence of the rider and the demands of the task. Symmetry, rhythm, coordination, stability and freedom of movement all give us evidence about whether the horse is able to do the work in comfort, and the rider is in symmetry with the horse’s movement to enable them to give clear signals.

4. Welfare outcomes

This is the longer view, because clear and effective communication should, over time, support good physical and mental welfare. Training that suits the individual horse develops their body and mind while reducing unnecessary stress or conflict, and building confidence. A horse in a lasting, harmonious relationship with their rider would be expected to show greater behavioural resilience, improved learning and a sustainable performance.

None of these components can work alone. Communication shapes behaviour, behaviour tells us how the horse is feeling, the body sets limits on what’s physically possible, and all of it feeds into horse welfare over time, which means a shortfall in one can’t be made up for by excellence in another. Spectacular movement doesn’t justify compromised welfare, and a calm looking horse isn’t proof of good welfare if the aids are inconsistent or the work is beyond what the horse can manage.

Figure 1. Four interconnected domains whose positive interactions give rise to what is termed harmony, and the points at which each domain can be observed or measured.

What we can already measure

A definition is only useful if it leads to something that can be seen, described and agreed on by more than one person. That is why measurement matters. It doesn’t replace the judgement of experienced horse people, but it gives that judgement something solid to work from, and several parts of what is termed ‘harmony’ can now be assessed in this way.

Rein tension is the clearest example, and it has been measured using strain gauges for around twenty years and is being constantly refined with measurement equipment readily available for use in training. When horses are given some choice over how much pressure they accept, they settle on a very light contact, with pressures recorded during ordinary riding often much higher than those horses would normally choose to avoid. Researchers have also looked at what happens when a horse opens its mouth or takes hold of the bit, and in both cases the pressure on the rein drops. This confirms that behaviour often described as an evasion is more likely to be the horse trying to avoid discomfort.

Other measures are constantly being developed. Pressure mats under the saddle, small wearable sensors and camera-based movement analysis allow rhythm, stride length, symmetry and rider stillness to be measured. Work on facial expression is newer but progressing quickly, with a coding system based on the facial muscles of the horse now established, and early computer models able to sort video footage into broad emotional categories with reasonable accuracy. Other approaches ask trained observers to score the overall behavioural expression of the horse against an agreed list of descriptors, which brings more consistency to the kind of whole horse impression that judges are already using.

Whilst much of these measurement approaches and technology have been developed for research purposes and as yet only used in controlled performance settings with small numbers of horses, they show a lot of potential and if used together, they can support expert judgement allowing consideration of whether a performance really shows the qualities the definition describes.

What would have to happen to objectively judge harmony

There is a strong case for judging harmony movement by movement rather than once at the end, and it could be made to work, but only alongside four other changes:

The sport has to agree what harmony means

What’s needed is a description of harmony that riders, trainers, officials and the watching public would all recognise, arrived at through proper consensus work rather than imposed from above.

Riders and trainers have to be able to learn it

A criterion is only useful if people can be taught how to produce it, which means describing harmony in terms a trainer can work with and a rider can recognise.

Judges have to be given the conditions to see it

Adding harmony to every movement will only improve on where we are now if the extra information comes from somewhere other than the judge’s view. Objective measures and analytical tools can provide that robust support.

The mark has to be able to change the result

At the moment a movement is scored largely on whether the horse performed the right figure in the right place, with very little room to record how calmly or otherwise it was done, so if the test is supposed to show harmony then gathering that information once at the end gives the score almost no influence over the final outcome.

So, can harmony be defined well enough to be scored?

The evidence suggests that it can, provided harmony is treated as more than an aesthetic impression. Describing it as the interaction between communication, the behaviour and affective state of the horse, biomechanical function and longer-term welfare outcomes makes it possible to define, teach and increasingly measure the qualities that sit underneath a successful horse and rider performance. Seen that way, harmony isn’t an indefinable artistic ideal but an evidence-informed feature of good horsemanship, and one that can be assessed fairly when science is there to support it.

The FEI rule revision proposals show a sport that is willing to look hard at how harmony is scored, and that is welcome. Whether collective marks are added or removed, judges will still need a shared idea of what they are looking at, and riders will still need to know what they are being asked to produce.

Selected References

Christensen JW, Munk R, Hawson L, Palme R, Larsen T, Egenvall A, König von Borstel UU, Rørvang MV (2021). Rider effects on horses’ conflict behaviour, rein tension, physiological measures and rideability scores. Applied Animal Behaviour Science 234, 105184. https://doi.org/10.1016/j.applanim.2020.105184

Christensen JW, Zharkikh TL, Antoine A, Malmkvist J (2011). Rein tension acceptance in young horses in a voluntary test situation. Equine Veterinary Journal 43(2), 223–228. https://doi.org/10.1111/j.2042-3306.2010.00151.x

Dyson S (2022). The Ridden Horse Pain Ethogram. Equine Veterinary Education 34(7), 372–380. https://doi.org/10.1111/eve.13468

Dyson S, Berger J, Ellis AD, Mullard J (2018). Development of an ethogram for a pain scoring system in ridden horses and its application to determine the presence of musculoskeletal pain. Journal of Veterinary Behavior 23, 47–57. https://doi.org/10.1016/j.jveb.2017.10.008

Dyson S, Pollard D (2020). Application of a Ridden Horse Pain Ethogram and its relationship with gait in a convenience sample of 60 riding horses. Animals 10(6), 1044. https://doi.org/10.3390/ani10061044

Eisersiö M, Rhodin M, Roepstorff L, Egenvall A (2015). Rein tension in 8 professional riders during regular training sessions. Journal of Veterinary Behavior 10(5), 419–426. https://doi.org/10.1016/j.jveb.2015.05.004

Eisersiö M, Yngvesson J, Hartmann E, Egenvall A (2023). Gaping for relief? Rein tension at onset and end of oral behaviors and head movements in unridden horses. Journal of Veterinary Behavior 59, 8–14. https://doi.org/10.1016/j.jveb.2022.11.009

Feighelstein M, Ricci-Bonot C, Hasan H, Weinberg H, Rettig T, Segal M, Distelfeld T, Shimshoni I, Mills DS, Zamansky A (2024). Automated recognition of emotional states of horses from facial expressions. PLoS ONE 19(7), e0302893. https://doi.org/10.1371/journal.pone.0302893

Gleerup KB, Forkman B, Lindegaard C, Andersen PH (2015). An equine pain face. Veterinary Anaesthesia and Analgesia 42(1), 103–114. https://doi.org/10.1111/vaa.12212

Hamilton KL, Lancaster BE, Hall C (2022). Equine conflict behaviors in dressage and their relationship to performance evaluation. Journal of Veterinary Behavior 55–56, 48–57. https://doi.org/10.1016/j.jveb.2022.07.011

Heidbuchel A, Van Rossom S, Molenaers N, Minguet P, Jonkers I (2023). Comparison of the effect of dressage rider skill level on physical fitness parameters and posture on an equestrian simulator. Journal of Equine Veterinary Science 121, 104187. https://doi.org/10.1016/j.jevs.2022.104187

Hintze S, Murphy E, Bachmann I, Wemelsfelder F, Würbel H (2017). Qualitative Behaviour Assessment of horses exposed to short-term emotional treatments. Applied Animal Behaviour Science 196, 44–51. https://doi.org/10.1016/j.applanim.2017.06.012

Kienapfel K, Link Y, König von Borstel U (2014). Prevalence of different head-neck positions in horses shown at dressage competitions and their relation to conflict behaviour and performance marks. PLoS ONE 9(8), e103140. https://doi.org/10.1371/journal.pone.0103140

Kienapfel K, Piccolo L, Cockburn M, Gmel A, Rueß D, Bachmann I (2024). Comparison of head–neck positions and conflict behaviour in ridden elite dressage horses between warm-up and competition. Applied Animal Behaviour Science 272, 106202. https://doi.org/10.1016/j.applanim.2024.106202

Minero M, Dalla Costa E, Dai F, Canali E, Barbieri S, Zanella A, Pascuzzo R, Wemelsfelder F (2018). Using qualitative behaviour assessment (QBA) to explore the emotional state of horses and its association with human-animal relationship. Applied Animal Behaviour Science 204, 53–59. https://doi.org/10.1016/j.applanim.2018.04.008

Minero M, Tosi MV, Canali E, Wemelsfelder F (2009). Quantitative and qualitative assessment of the response of foals to the presence of an unfamiliar human. Applied Animal Behaviour Science 116(1), 74–81. https://doi.org/10.1016/j.applanim.2008.07.001

Napolitano F, De Rosa G, Braghieri A, Grasso F, Bordi A, Wemelsfelder F (2008). The qualitative assessment of responsiveness to environmental challenge in horses and ponies. Applied Animal Behaviour Science 109(2–4), 342–354. https://doi.org/10.1016/j.applanim.2007.03.009

Piccolo L, Kienapfel K (2019). Voluntary rein tension in horses when moving unridden in a dressage frame compared with ridden tests of the same horses. A pilot study. Animals 9(6), 321. https://doi.org/10.3390/ani9060321

Wathan J, Burrows AM, Waller BM, McComb K (2015). EquiFACS: the equine facial action coding system. PLoS ONE 10(8), e0131738. https://doi.org/10.1371/journal.pone.0131738

Wolframm I (2023). Let them be the judge of that: bias cascade in elite dressage judging. Animals 13(17), 2797. https://doi.org/10.3390/ani13172797