The Veterinary Journal 181 (2009) 5–11

Contents lists available at ScienceDirect

The Veterinary Journal

journal homepage: www.elsevier .com/ locate/ tv j l

Review

How equitation science can elucidate and refine horsemanship techniques

Deborah Goodwin a,*, Paul McGreevy b, Natalie Waran c, Andrew McLean d

a School of Psychology, University of Southampton, Southampton, SO17 1BJ, UKb Faculty of Veterinary Science (B19), University of Sydney, NSW 2006, Australiac Animal Welfare Group, Unitec New Zealand, Private Bag 92025, Auckland, New Zealandd Australian Equine Behaviour Centre, Clonbinane Road, Broadford, Vic 3658, Australia

a r t i c l e i n f o

Keywords:HorseTrainingEquitationLearning theoryEthologyCognition

1090-0233/$ - see front matter � 2009 Elsevier Ltd. Adoi:10.1016/j.tvjl.2009.03.023

* Corresponding author. Tel.: +44 23 8059 5000.E-mail address: d.goodwin@soton.ac.uk (D. Goodw

a b s t r a c t

The long-held belief that human dominance and equine submission are key to successful training andthat the horse must be taught to ‘respect’ the trainer infers that force is often used during training. Manyhorses respond by trialling unwelcome evasions, resistances and flight responses, which readily becomeestablished. When unable to cope with problem behaviours, some handlers in the past might have beenencouraged to use harsh methods or devices while others may have called in a so-called ‘good horseman’or ‘horse whisperer’ to remediate the horse. Frequently, the approaches such practitioners offer could notbe applied by the horse’s owner or trainer because of their lack of understanding or inability to apply thetechniques. Often it seemed that these ‘horse–people’ had magical ways with horses (e.g., they only hadto whisper to them) that achieved impressive results although they had little motivation to divulge theirtechniques.

As we begin to appreciate how to communicate with horses sensitively and consistently, misunder-standings and misinterpretations by horse and trainer should become less common. Recent studies havebegun to reveal what comprises the simplest, most humane and most effective mechanisms in horsetraining and these advances are being matched by greater sharing of knowledge among practitioners.Indeed, various practitioners of what is referred to here as ‘natural horsemanship’ now use techniquessimilar to the ‘whisperers’ of old, but they are more open about their methods. Reputable horse trainersusing natural horsemanship approaches are talented observers of horse behaviour and respond consis-tently and swiftly to the horse’s subtle cues during training. For example, in the roundpen these trainersapply an aversive stimulus to prompt a flight response and then, when the horse slows down, movestoward them, or offers space-reducing affiliative signals, the trainer immediately modifies his/her agonis-tic signals, thus negatively reinforcing the desired response.

Learning theory and equine ethology, the fundamentals of the emerging discipline of equitation sci-ence, can be used to explain almost all the behaviour modification that goes on in these contexts andin conventional horsemanship. By measuring and evaluating what works and what does not, equitationscience has the potential to have a unifying effect on traditional practices and developing branches ofequitation.

� 2009 Elsevier Ltd. All rights reserved.

Introduction

The word ‘horsemanship’ is often used to describe the art of rid-ing, driving, handling and managing horses. It is considered a skillacquired through knowledge of and experience with horses. Tradi-tional horsemen use gradual habituation to new equipment andexperiences to train horses. There is also an element of good horse-manship that is frequently described as ‘natural’ or innate; in otherwords, some horsemen are thought to be ‘born with the ability’ towork well with horses (e.g., Blackfoot Native Americans) (Ewers,

ll rights reserved.

in).

1955), English and Scottish farriers, carters and ostlers (Evans,1960), horse professionals from European, Asian and North Amer-ican cultures (van Dierendonck and Goodwin, 2005). This mani-fests as a familiarity with the horses’ inherent postural andlocomotory movements.

Such horse sense can appear innate, but in reality it is likely tobe a combination of natural ability combined with learning thathas developed with time and experience. It is characterised bysmooth body movements, calmness and patience. Many horsemenacknowledge that tactile and visual cues may be more salient tohorses than auditory and olfactory signals, although they acceptthat these, too, play a role in communication within and betweenspecies. So, given the relatively minimal importance of vocalisation

6 D. Goodwin et al. / The Veterinary Journal 181 (2009) 5–11

in the equid ethogram, it is not surprising that horse sense is fur-ther typified by quietness in humans (McDonnell, 2003).

It may be helpful at this point to define some of the terms usedin learning theory. Positive reinforcement is the addition of some-thing pleasant (a reinforcer) to reward the desired response; nega-tive reinforcement is the subtraction of aversive stimuli (e.g.,pressure) to promote the desired response; positive punishment isthe addition of aversive stimuli to reduce an unwanted responseand negative punishment the removal of a reward to reduce an un-wanted response.

According to Skinner (1938) and Nicol (2005), animals learn byusing information about their environment to tailor their re-sponses to changes. Training institutes novel responses by rein-forcing desirable behaviours and suppressing undesirable ones(McGreevy and Boakes, 2007). The two processes are fundamentalto effective horse training in order to achieve a mount that is reli-ably under stimulus control originating from the rider, rather thanone that is responding primarily to environmental stimuli.

Fig. 1. The curb bit works by squeezing the mandible as a result of pressure exertedon the tongue, diastema and the curb groove (photograph by Minnas Finland,reproduced with permission).

Historical and cultural influences on horsemanship

According to Greek mythology, the first ‘horseman’ was the god-dess Minerva, who invented the bridle to help Bellerophon controlPegasus, the winged horse. Archaeological evidence indicates thathorses were ridden in the Ukraine some 6000 years ago (Telegin,1986), which is even earlier than when they were first used asdraught animals in North Africa (Clutton-Brock, 1999). The mainevidence of human intervention is from the skulls of horses inwhich tooth wear is thought to be comparable with that found inhorses ridden in a metal ‘bit’ (a mouthpiece used to control speedand direction). Such bits have been used throughout history andprogressed from twisted bronze bars around 1400 BC, to jointedbits with the familiar ‘nutcracker’ action around 1000 BC; the mod-ern jointed snaffle is similar to these.

As the uses for horses diversified in different cultures, bit de-signs proliferated and developed into the variety of bits now usedby modern riders. This is testament to the multiplicity of attemptsby riders to communicate effectively with the ridden horse, and isperhaps also evidence of their failure to achieve this in many situ-ations. Horses under stimulus control should respond to light reinsignals delivered via bit pressure on the tongue and bars of themouth (e.g., a ‘French link’ snaffle) or in some cases including pres-sure on the poll and chin groove (e.g., curb bits, see Fig. 1), makingthe need for often progressively harsher methods (including bits)redundant. It is important to note that curb bits are deceiving inthat the rider feels only very light pressure but the horse may feelalmost intolerable pain which is why they work but if only the trai-ner has trained the horse to slow from lighter aids using negativereinforcement correctly.

An important initial motivation for riding on the back of a horseseems to have been the advantages gained in war. For example,well before 1000 BC, the Persians were using horses and chariotseffectively in wars against the Greeks, which encouraged theGreeks to develop their own cavalry. It is clear that the use ofhorses was often a pivotal factor in the outcome of conflicts at thistime (Anderson, 1961). While there is no doubt that early horse-men were apparently skilled, their training methods may now ap-pear unsophisticated and harsh. For example, Persian horsetrainers strongly advocated using hobbles to restrain horses andprevent them from straying, and muzzles to prevent them frombiting.

The earliest known text on horse training was written byKikkuli in 1400 BC in the language of the Hittite region, which to-day stretches across South-Eastern Turkey, Northern Syria andNorthern Iraq. Kikkuli was a master horse trainer and wrote mainly

about conditioning (fittening) Hittite war horses through exerciseand feeding, rather than about the type of early handling or train-ing used for breaking-in horses (Waran et al., 2002). In 350 BC,Xenophon, an enlightened Greek cavalry officer and philosopher,wrote The Art of Horsemanship (translated by Morgan, 1979), inwhich he advocated what he considered to be humane methodsof training, handling and managing horses. For example, he recom-mended that young cavalry horses should be trained by profes-sional trainers, advised that foals should experience kind handlingbefore being trained, and suggested that owners should keep a closeeye on how their mounts were trained.

While Xenophon may have lacked specific knowledge of learn-ing theory, he recognised the importance of a good human–horserelationship (for example, he stated that during the process ofhorse training, the horse should associate being alone with beinghungry, thirsty or annoyed by flies, and the presence of a man withfood, water and relief from flies). He went on to write that a horseshould be rewarded with kindness for doing what the trainerwanted. However, he did advocate creating hyper-reactive states,perhaps similar to those in modern dressage or show-jumpinghorses to make them look proud, high mettled, and fiery by apply-ing rein pressure and acceleration (leg pressure) signals concur-rently. This practice conflicts with the approach now advocatedthrough equitation science, where conflicting leg and rein signalsare considered to confuse the horse and, in time, lead to dulledresponses.

Several centuries ago in Britain, a secret society of horsemen(The Society of the Horseman’s Word) came into existence andseems to have been somewhat akin to Freemasonry. The society

D. Goodwin et al. / The Veterinary Journal 181 (2009) 5–11 7

probably had its origins in the then-common Romano-British craftor trade societies. Membership was open to men working withhorses (such as ploughmen, farriers, ostlers and carters) but ex-cluded farmers and women. Following quasi-occult initiation cere-monies, secrets of the trade were shared, including knowledge ofmanaging and working with horses, and finally, the ‘Horseman’sWord’ would be revealed to the new member. This word becameassociated with magical powers over horses when whispered intotheir ears, but it is likely to have been more of a red herring used todistract the attention of human observers from the handling tech-niques and secret recipes used by the trainers. These horse whis-perers used their knowledge of dietary additives and the horse’ssense of olfaction to control and modify their behaviour (Evans,1960) perhaps reflecting the earlier practices of the Hittite Kikkuli.They also used consistent triggers from their own behavioural rep-ertoire to cue responses with great clarity and reward appropriateresponses with exquisite timing. Most likely, the term ‘horse whis-perer’ was originally used for members of this society, but was andstill is, adopted by horse trainers wishing to suggest that they havemysterious powers over horses.

Historically, horse whisperers were called upon to deal with fre-quently occurring problems between people and their horses.Since brute force was often used in traditional horse training inwestern cultures to achieve compliance, it is probable that manyhorses would have learned unwelcome evasions, resistances andflight responses. It was also commonly believed that human dom-inance and equine submission were critical to success and thedevelopment of so-called ‘respect’ on the horse’s part, which canbe a recipe for creating problem behaviours. The whisperers some-times achieved impressive results with difficult animals, but didnot divulge the techniques they used. Their deftness and the ab-sence of any apparent vocalisation supported the belief that theyneeded only to whisper to horses to produce the desired results.It added to their mystique to allow the uninitiated to believe thatwhispering, rather than training or other management techniques,was the key.

Throughout history numerous horsemen have demonstratedtheir ability to exploit the marketplace and teach similar methodsfor commercial reward. Unfortunately, not all followers of thesemethods are as effective as the originating trainers. Poor techniquecan lead to disappointing results and eventually frustration inpractitioners, which may result in abuse, confusion and conflictbehaviours in horses, which may then contribute to human acci-dent and equine wastage statistics.

Natural horsemanship and the natural behaviour of the horse

With the proliferation of horses as companion animals over thepast 25 years, a range of methods and gadgets (e.g., natural horse-manship rope halters) for training or re-training horses and/or rid-ers has emerged. These methods are often influenced more bycultural attitudes to animal welfare and utility than by efficacyand humaneness. Yet despite the long history of horsemanship, itappears that we have progressed little in our training techniques.Given that humans are now accustomed to using technology,including mobile phones, tube trains, automobiles, and other formsof communication and/or transport, this move away from dailycontact with equids on the mass human scale may have beeninstrumental in the ‘non-progress’ in training equids.

Most current techniques are still based on the traditional meth-ods and equipment used by the Greeks and Romans. What are ac-cepted as conventional or traditional horse-training methods alsovary between different cultures and traditions (e.g., between Wes-tern Europe and Kazakhstan). Even those training methods pro-moted as more innovative or sympathetic, and referred to here

as natural horsemanship, are not necessarily new, and can be dif-ficult to apply in a practical context except by an experienced trai-ner. Good natural horsemanship trainers are talented observers ofhorse behaviour who respond precisely to subtle cues during train-ing. Some techniques, when subjected to closer scrutiny (e.g.,roundpen work), could be considered aversive (Krueger, 2007)when used by less-effective trainers. Personal opinions have oftenbeen presented as facts because there has been little objectivestudy of horse behaviour in the natural or domestic environment.Natural horsemanship appears to focus on establishing communi-cation between man and horse, but some effects would be ex-plained more scientifically as adaptive flight responses andlearned responses to stimuli.

A good example of this is seen in the use of the roundpen fortraining young horses. The system is an adaptation of the Gyrus,used by the Romans, which in turn was borrowed from the Greeks,who started training young horses in small, enclosed circular are-nas filled with deep sand or shavings. These arenas had high, solidsides so that the horse’s attention was not distracted from its han-dler (Waran et al., 2002), and so were quite different to many of theopen-sided roundpens in use today. With the roundpen trainingapproach used today by well-known horsemen (see, for example,Roberts, 2000; Parelli, 2003), the circular shape enables the handlerto keep the horse moving during training. This helps to ensure thatthe horse rapidly learns to respond to the trainer’s signals, to ap-proach the trainer and not to move away unless prompted to doso. This in turn allows the trainer to approach the horse and grad-ually habituate it to being handled and restrained.

Roundpen work has been described as a method based on anelement of the ethogram of the horse in which a dominant memberof a group will chase another to become its leader, force its compli-ance or win its respect. However, the theory lacks credibility be-cause, apart from anecdotal accounts, these aspects of equinesocial behaviour have never been recorded. Published ethologicalstudies in the US and UK show that leaders in horse society arerarely the most aggressive or dominant individuals (Tyler, 1972;Waring, 2002), that horse society is matriarchal, based on long-term associations between related individuals and preferred com-panions, and that the usual response to aggression is avoidance(Fraser, 1992).

The merits of hands-off roundpen training are proposed as hu-mane and non-coercive, but both unhandled and handled horseswill try to avoid being chased. It is likely that horses make associ-ations between their fearful states and the presence of the aggres-sor in the middle of the roundpen (see Fig. 2). In addition, fear andstress are known to reduce effective learning (Moberg and Mench,2000) and could therefore be considered counter-productive if thetraining is to be effective.

It is beguiling to think that we can learn to ‘speak horse’ and en-ter their social hierarchy by mimicking their signals and behaviour.However, we are unable to mimic their signals with any subtlety aswe do not have the same visual signalling structures (e.g., ears thatmove; see also McGreevy et al., 2009). For the mounted rider, com-munication difficulties escalate, because any ability humans mayhave to relate to horses using equine communication cues is ofnegligible value once they mount. There is no evidence that horsesperceive us as ‘honorary horses’, or that we can insert ourselvesinto their social organisation. The aim in roundpen work is to im-pose the human will on the horse by understanding its behaviour,even exerting dominance over it, until the horse learns to respectour leadership. Although attractive, unfortunately these ideas areinherently flawed and potentially problematic. For example,regardless of the interpretations placed on the relationshipbetween a horse and its handler, any roundpen activity that allowsa horse to flee in response to the presence of a human can becounter-productive since it can allow a horse to add flight to its

Fig. 2. The danger of roundpen work is that horses can make associations betweentheir fearful states and the presence of the aggressor in the middle of the roundpen(photograph by Carol Wilcocks, reproduced with permission).

8 D. Goodwin et al. / The Veterinary Journal 181 (2009) 5–11

repertoire of responses to humans. This response is unhelpful in allworking horse contexts.

The usual explanation of roundpen work in natural horseman-ship is based on the premise that a herd is organised accordingto social status and ritualised conflict, which academic ethologistshave shown is not true (see, for example Waring, 2002; Fraser,1992). Such beliefs can cause major problems, for example whenhorses trial aggression in a bid to switch off human chasing behav-iours in a roundpen. We also have some evidence of horses’ abilityto understand categories (e.g., shape, colour and size) (Visser et al.,2003), but not yet that they have the cognitive capacity to under-stand concepts such as ‘trust’, ‘respect’ and ‘leadership’.

Equitation science and horse training

Recent studies in the use of applied learning theory have begunto expose what are the simplest, most humane and most effectivemechanisms in horse training. As we appreciate how to use dis-crete signals to communicate with horses sensitively and consis-tently, misunderstandings and misinterpretations by both horseand human will decrease. These advances are being matched bya greater sharing of knowledge among practitioners. As an exam-ple, natural horsemanship trainers can be appreciated as talentedobservers of horse behaviour who respond consistently and swiftlyto subtle cues during training. They apply aversive stimuli toprompt a flight response and then, when the horse slows downor moves toward them or offers space-reducing affiliative signals,the trainer removes the aversive stimuli and thus negatively rein-forces the desired response, without necessarily understanding thetheory behind their actions. When subjected to scientific enquiry,the strengths and weaknesses in their methods can be identifiedand improved with the shared ultimate aim of improving welfareand reducing wastage.

In another example, the human in the centre of the roundpenmay be positively reinforcing the horse’s approaches by modifyinghis/her agonistic signals and possibly by grooming the animal’sforehead. However, Krueger’s study (2007) of roundpen trainingsuggests that grooming in this area is not as reinforcing as somenatural horsemanship practitioners seem to assume. According toFeh and de Mazières’ study (1993) of horses’ preferred groomingsites, the forehead was rarely groomed by conspecifics and groom-ing on the withers was preferred.

The horse’s approach behaviour in the roundpen is shapedthrough progressive improvements. However, if the horse stops

approaching, the human punishes it by immediately chasing itaway until it actively approaches again. Join-up (characterised bythe horse following the human) (Roberts, 2000) can be establishedbecause it effectively utilises learning theory and, in particular,aversive conditioning. However, Krueger (2007) reported that afterjoin-up had been demonstrated in the roundpen, horses did notfollow the trainer when outside the roundpen; in other words,the following response was not generalised. This is not surprisingas the join-up technique does not place the response to approachhumans under stimulus control. Krueger (2007) concluded thatin the roundpen horses simply learned to avoid being chased. Whatoccurs in a roundpen can be explained by a continued analysis oflearning theory and equine cognitive ethology (i.e., a study of thethought processes and mental capabilities of horses) rather thanimplying that a relationship has been brokered by the process ofjoin-up.

Understanding horse behaviour

Ethology allows us to understand the natural and adaptivebehaviour of horses, but horses did not evolve to be ridden. Sincedomestication we have bred horses for specific purposes, but whenwe train and ride them we also need to use learning theory devel-oped in psychology to understand learned associations betweenstimulus and response (Warren-Smith and McGreevy, 2008). Addi-tionally, we need input from a range of diverse disciplines, includ-ing veterinary science, sports physiology, nutrition, biomechanicsand engineering. It is important that we use learning theory to ex-plain the behavioural modification that goes on in any trainingcontext, because relying on anthropocentric or anthropomorphicexplanations, however alluring and plausible, adds an unnecessarylayer of complexity and leaves some aspects of the horse’s motiva-tion open to interpretation. This generally leads to confusion andsometimes to compromised welfare for the horse.

The commercial success of the natural horsemanship move-ment has led to a questioning of some traditional practices, hownovel techniques operate and how the terms relating to horsetraining and riding also relate to what is known through psychol-ogy, ethology and veterinary science. The burgeoning demand forbest practice in horse handling and training (a practice that isunderpinned by empirical data rather than belief systems) hasprompted the emergence of a fascinating new area of scientific re-search (Minero et al., 2006; Goodwin et al., 2007; McGreevy andMcLean, 2007). Equitation science does not seek to turn equitationinto a science per se but aims to develop scientific methods tostudy, measure and interpret interactions between horse and riderduring equitation. Measuring these variables is important becauseit allows techniques to be compared and assessed in order to dem-onstrate what works and what does not. It also allows us to mea-sure the welfare consequences of doing the wrong thing.Equitation scientists acknowledge the contribution of naturalhorsemanship in advancing equestrian techniques, but continueto observe and study equine behaviour and training in the searchfor further improvements.

Roles for science in horse welfare

The extent to which sport horses are coerced to perform is oftenthe focus of welfare debates. For example, the traditional speedand endurance phases (roads-and-tracks and steeplechase) havedisappeared from the sport of horse-trials. But with the numberof fatalities within eventing at an all-time high, it is debatablewhether the removal of phases A and B has made it easier forcompetitors to get to phase D and therefore allows less well-prepared combinations to tackle cross-country courses. In racing,

D. Goodwin et al. / The Veterinary Journal 181 (2009) 5–11 9

steeplechasing and the use of the whip may be moderated as a re-sult of pressure from groups outside the racing industry. Thehorse-welfare lobby has also begun to voice its concerns aboutother elite equestrian events. For example, some riders and train-ers consider the dressage movement called piaffe (in essence,trotting on the spot) unnatural and therefore unacceptable. Theuse of hyperflexion in training is now a topic of heated debateamong welfare organisations, riders, sports organisations and thepublic. In this technique, which is sometimes referred to as long-deep-and-round training or Rollkur, the horse is trained to carryits head low with its cervical vertebrae maximally flexed (chinclose to the chest; Fig. 3) in the belief that the hindquarters arethen engaged and that the activity and power of the hindlegs areimproved (McGreevy et al., 2005). One of the drawbacks of thisfocus is that it ignores other potentially problematic ridingtechniques, such as high head carriages with tight reins.

Equitation science will play a vital role in influencing the out-come of these debates. In particular, in dressage competition,emergent technology will remove subjectivity from judging andwill underpin the emergence of high-welfare dressage. Calibratedrein-tension gauges and pressure-sensitive pads can measure thestrength and frequency of a rider’s signals to the horse. Ridersand trainers can integrate technology into their training methodsto measure contact and lightness objectively. Science may be ableto help us value training of any manoeuvre that depends on and isachieved through lightness (i.e., attesting to self-carriage and thehorse’s self-maintenance of rhythm, straightness and outline,McGreevy, 2007). This raises the possibility that, in the future,dressage scores could be awarded on an objective assessment ofthe most humane riding and training techniques.

Equitation science needs to address the possibility of learnedhelplessness occurring in horses used in sport and leisure (Hallet al., 2008) and as a consequence of any radical restraining tech-nique. This will provide physiological measures of stresses thataccompany passive coping. In addition, horse welfare will improveif we can identify horses that are simply not capable of a response.For example, an ‘on-board’ device that measures physiological orresponse fatigue in real time may one day allow riders to knowwhen to quit before inducing a cycle of over-training. The same ap-proach may allow judges and veterinarians to monitor the welfareof horses in competition. This sort of advance has the potential toensure that fatigued horses are not pushed or beaten in the nameof sport, or any other activity.

The racing industry and other performance sports may becomemore engaged with equitation science as the physiology of thehorse at exercise and cognitive aspects that go beyond learningtheory alone become further elucidated (Goodwin, 2007). The sig-nificance of subjective states, such as mood, trust, stoicism andeven attitude, may then vie with cardiac output and gaseous ex-

Fig. 3. A hyperflexed horse (photograph by Julie Taylor, reproduced withpermission).

change as behavioural qualities to be measured and selected for.To help explain racing dynamics, there needs to be more scrutinyof the (currently anecdotal) evidence for communication amonghorses competing as a group.

The way forward

There is still much to discover about how horses learn from theenvironment, from each other and from human interactions (Mur-phy and Arkins, 2007). For example, context-specific learningneeds further investigation (e.g., what are the effects of taking ahorse to a show, or of the direction in which a horse approachesan obstacle?). Furthermore, it is suggested that horses somehowpercolate training and come back to work after a break with evi-dence of having made salutary advances in their learning, almostas if the training needs to be interrupted ‘to mature’ as a resultof neuronal change, growth and vascularisation. It would be inter-esting to investigate more how training contingencies do matureor decline with gaps in training. For example, horses could be eval-uated on a weekly basis to show how much learning is retainedfrom one week to the next.

There are many questions remaining about positive reinforce-ment. Apart from the importance of timing, the quality of deliveredrewards during horse training has yet to be fully explored, espe-cially for in-hand work. We should never assume that the resultsto date (e.g., Warren-Smith and McGreevy, 2007) represent thecomplete picture. For example, if we use latency to complete amaze to measure the success of training horses with positive rein-forcement, we may overlook the possibility that experimentalequine subjects that sense food arriving in their mouths may nat-urally slow the movement of their legs (as they would at pasture).

Training and performance differences between stereotypichorses (such as crib-biters and weavers) and other horses meritthorough longitudinal investigation so that the long-term effectsof stereotypy on learning can be evaluated (Hausberger et al.,2007; Parker et al., 2008a). This may help to create further incen-tives for breeders and trainers to avoid management styles thattrigger the emergence of stereotypies in predisposed horses(Parker et al., 2008b). While ongoing research efforts are likely toidentify specific risk factors for all equine stereotypies, they shouldbe aimed at refining stable management rather than at selectinghorses that will tolerate existing practices (Thorne et al., 2005).When developing their stereotypies, crib-biters and weavers canbe regarded as useful sentinels of sub-optimal husbandry, sentinelsthat should alert us to problems that are being encountered by allresident horses, not just those that produce a stereotypic responseto the challenge.

Temperament characteristics of elite horses should be measuredand characterised more fully. This will promote our understandingof those breed characteristics that favour certain equestrian activi-ties and allow us to select for certain traits, where desirable andpossible. An ongoing exploration of temperament (Visser et al.,2003) and left–right motor bias (McGreevy and Rogers, 2005;McGreevy and Thomson, 2006) as predictors of reactivity and suit-ability for certain sports and work may reduce behavioural wast-age. However, it is important to retain a long-term perspective inany breeding program based on any suite of tests for reactivity ortemperament so that the effects of selecting for one set of traitscan be tested in case it inadvertently brings along any unwantedbaggage (i.e., undesirable or unexpected traits).

Training and re-training

Despite the enormous variety in horse sports and equine work,only four basic responses are required for them all: stop, go, turn

10 D. Goodwin et al. / The Veterinary Journal 181 (2009) 5–11

the forequarters, and turn the hindquarters (McLean, 2003; seeFig. 4). A solid foundation training of these basic responses allowsa horse to be trained with a customised plan specific to the disci-pline it is intended for, but it also means that the horse can bere-trained later for secondary work after its first career is over(McLean and McLean, 2008).

Training relies on timing and consistency and so, by deduction,de-training soon arises when there is inconsistency and variabletiming. Enlightened trainers are consistent with their signals andthe way in which they set up puzzles for the animal to solve (i.e.,the way in which they pose questions). Good technique that re-flects these qualities is more sustainable than the size or strengthof handlers or, indeed, any devices that allow handlers to applymore force.

Our greatest responsibility is to ensure we never forget that thehorse’s welfare is paramount. Therefore, every horse trainer shouldmaintain an open mind about possible limitations in horse learningand confusion arising from (less-than-optimal) training methods.This is especially important when the vast range of required re-sponses in the trained horse is compared with the limited numberof sites on the animal’s body for eliciting those responses. If we aredealing with an animal whose ability to extrapolate is currentlyunknown, we must always be mindful of the potentially confusingeffects of applying pressure signals in common or overlapping siteson its body to elicit different responses.

Recent research in several countries has revealed similar pat-terns, and it is worrying to consider that much tragic behavioural

Fig. 4. Training the horse to stop, go, turn the forelegs and turn the hindlegs is thebasis for all movements required in equitation.

wastage (Ödberg and Bouissou, 1999) is a result of our unclearinteractions and the impossible expectations humans place ontheir horses. What is needed is a reappraisal and restructuring ofcontemporary horse training within the framework of establishedand empirically tested principles of learning, and that is offeredonly by equitation science.

Conclusions

Traditional training methods have been goal-driven, with themain objective being to produce a safe and effective animal for aparticular task. In scientific terms, the process of training can be de-fined as the intentional modification of the frequency and/or inten-sity of specific behavioural responses. These can be achievedthrough different approaches, but generally by the shaping of a de-sired response, where the specific behaviour is positively or nega-tively reinforced in a step-wise fashion toward a specific end point,and undesirable responses suppressed using negative reinforce-ment and/or positive punishment.

Traditional horsemen use gradual habituation to new equip-ment and experiences to train horses. Habituation is combinedwith operant conditioning techniques that rely on negative rein-forcement, where the horse learns to perform the desired behav-iour through moving away from pressure or avoiding an aversivestimulus. When carried out in a structured way by a knowledge-able trainer, all the training and various techniques that can beused should involve as little stress to the horse as possible. How-ever, learning new things, getting things wrong and being exposedto new situations will almost inevitably cause the animal some le-vel of discomfort.

Many traditional practices in riding and stable managementpush horses beyond their natural adaptive behavioural repertoire;some horses cope and others do not. Equitation presents significantethological challenges and, in many cases, training fails to reflectadequately the physical abilities and learning capacity of the horse.All too often the source of a rider’s difficulty emerges from aninability to understand his own role in confusing the horse by pro-viding contradictory signals. We must also recognise that mosthorses are extremely tolerant, but just because we can train a horseto do something that does not make it an ethically sound practice.After all, dogs have been trained to salivate in response to electricshocks, but we would never assume that they enjoyed theexperience.

Equitation science argues that successful and humane trainingrelies on the horse trainer having an understanding of the applica-tion of learning theory, horse ethology, physiology and biomechan-ics, as well as recognising the goals and limitations of the trainingapproach being used. Effective and humane training relies on suchknowledge, as does the ability to deal with any problem behaviourthat may arise when training techniques sometimes break down.

Conflict of interest statement

None of the authors of this paper has a financial or personalrelationship with other people or organisations that could inappro-priately influence or bias the content of the paper.

References

Anderson, J.K., 1961. Ancient Greek Horsemanship. University of California Press,Berkeley, USA.

Clutton-Brock, J., 1999. A Natural History of Domesticated Mammals. CambridgeUniversity Press, Cambridge.

Evans, G.E., 1960. The Horse in the Furrow. Faber and Faber, London.Ewers, J.C., 1955. The horse in Blackfoot Indian culture. Smithsonian Institute

Bureau of American Ethnology Bulletin 159, 33–47.

D. Goodwin et al. / The Veterinary Journal 181 (2009) 5–11 11

Feh, C., de Mazières, J., 1993. Grooming at a preferred site reduces heart rate inhorses. Animal Behaviour 46, 1191–1194.

Fraser, A.F., 1992. The Behaviour of the Horse. CAB International, Wallingford, Oxon,UK.

Goodwin, D., 2007. Equine learning behaviour: what we know, what we don’t knowand priorities for future research. Behavioural Processes 76, 17–19.

Goodwin, D., Heleski, C., McGreevy, P., McLean, A., Randle, H., Skelley, C., vanDierendonk, M., Waran, N. (Eds.), 2007. Proceedings of the 3rd InternationalEquitation Science Symposium. Michigan State University Press, Michigan, USA.

Hall, C., Goodwin, D., Heleski, C., Randle, H., Waran, N., 2008. Is there evidence oflearned helplessness in horses? Journal of Applied Animal Welfare Science 11,249–266.

Hausberger, M., Gautier, E., Muller, C., Jego, P., 2007. Lower learning abilities instereotypic horses. Applied Animal Behaviour Science 107, 299–306.

Krueger, K., 2007. Behaviour of horses in the ‘round-pen technique’. Applied AnimalBehaviour Science 104, 162–170.

McDonnell, S.M., 2003. The equid ethogram. A Practical Field Guide to HorseBehavior. The Blood Horse Inc., Lexington, Kentucky.

McGreevy, P.D., 2007. The advent of equitation science. The Veterinary Journal 174,492–500.

McGreevy, P.D., Boakes, R.A., 2007. Carrots and Sticks: Principles of Animal Training.Cambridge University Press, Cambridge.

McGreevy, P.D., McLean, A.N., 2007. The roles of learning theory and ethology inequitation. Journal of Veterinary Behavior: Clinical Applications and Research 2,108–118.

McGreevy, P.D., Rogers, L.J., 2005. Motor and sensory laterality in thoroughbredhorses. Applied Animal Behaviour Science 92, 337–352.

McGreevy, P.D., Thomson, P.C., 2006. Differences in motor laterality in breeds ofperformance horse. Applied Animal Behaviour Science 99, 183–190.

McGreevy, P., McLean, A., Warren-Smith, A., Goodwin, D., Waran, N., 2005. In:Proceedings of the 1st International Equitation Science Symposium. PostGraduate Foundation, University of Sydney, Melbourne, Australia.

McGreevy, P.D., Oddie, C., Burton, F.L., McLean, A.N., 2009. The horse–human dyad:can we align horse training and handling activities with the equid socialethogram? The Veterinary Journal 181, 12–18.

McLean, A.N., 2003. The Truth About Horses. Penguin, Melbourne, Australia.McLean, A.N., Mclean, M.M., 2008. Academic Horse Training: Equitation Science in

Practice. Australian Equine Behaviour Centre, Victoria, Australia.Minero, M., Canali, E., Warren-Smith, A., McLean, A., Goodwin, D., Zetterqvist, M.,

Waran, N., McGreevy, P., 2006. In: Proceedings of the 2nd InternationalEquitation Science Symposium. Fondazione Iniziative Zooprofilattiche, Milano,Italy.

Moberg, G.P., Mench, J.A., 2000. The Biology of Animal Stress: Basic Principles andImplications for Animal Welfare. CAB International, Wallingford, Oxon, UK.

Murphy, J., Arkins, S., 2007. Equine learning behaviour. Behavioural Processes 76,1–13.

Nicol, C., 2005. Learning abilities of the horse. In: Mills, D.S., McDonnell, S.M. (Eds.),The Domestic Horse: The Origins, Development and Management of Behaviour.Cambridge University Press, pp. 169–183.

Ödberg, F.O., Bouissou, M.F., 1999. The development of equestrianism from thebaroque period to the present day and its consequences for the welfare ofhorses. Equine Veterinary Journal 28, 26–30.

Parelli, P., 2003. Natural horse-man-ship. Western Horseman. Colorado Springs,USA.

Parker, M., Redhead, E., Goodwin, D., McBride, S., 2008a. Impaired instrumentalchoice in crib-biting horses (Equus caballus). Behavioural Brain Research 191,137–140.

Parker, M., Redhead, E., Goodwin, D., 2008b. Survey of breeders’ management ofhorses in Europe, North America and Australia: comparison of factorsassociated with the development of abnormal behaviour. Applied AnimalBehaviour Science 114, 206–215.

Roberts, M., 2000. Join-Up – Horse Sense for People. Harper Collins, London.Skinner, B.F., 1938. The Behavior of Organisms. Appleton-Century, New York, USA.Telegin, D.Y., 1986. Dereivka, a Settlement and Cemetery of Copper Age Horse

Keepers on the Middle Dneiper. BAR International Series. p. 287.Thorne, J.B., Goodwin, D., Kennedy, M.J., Davidson, H.P.B., Harris, P., 2005. Foraging

enrichment for individually housed horses: practicality and effects onbehaviour. Applied Animal Behaviour Science 94, 149–164.

Tyler, S., 1972. The behavioural and social organisation of new forest ponies. AnimalBehaviour Monographs 5, 85–196.

van Dierendonck, M., Goodwin, D., 2005. Social contact in horses: implications forhuman–animal relationships. In: de Jonge, F., van den Bos, R., Zwart, H.,Lijmbach, S. (Eds.), Animals in Science and Philosophy Series: Human AnimalRelationships. van Gorcum BV, Assen, pp. 65–81.

Visser, E.K., Reenen, C.G., van Schilder, M.B.H., Barneveld, A., Blokhuis, H.J., 2003.Learning performances in young horses using two different learning tests.Applied Animal Behaviour Science 80, 311–326.

Waran, N.K., McGreevy, P., Casey, R.A., 2002. Training methods and horse welfare.In: Waran, N.K. (Ed.), The Welfare of Horses. Kluwer Academic Publishers,Dordrecht, The Netherlands.

Waring, G.H., 2002. Horse Behavior, second ed. William Andrew Publishing,Norwich, NY.

Warren-Smith, A.K., McGreevy, P.D., 2007. The use of blended positive and negativereinforcement in shaping the halt response of horses (Equus caballus). AnimalWelfare 16, 481–488.

Warren-Smith, A.K., McGreevy, P.D., 2008. Equestrian coaches’ understanding andapplication of learning theory in horse training. Anthrozöos 21, 153–162.

Xenophon, 1979. The Art of Horsemanship. J.A. Allen, UK (translated Morgan, N.).