Since Monty Roberts wrote The Man Who Listens to Horses in 1999, horse enthusiasts have been inundated with talk of round pens, respect, joining-up, etc. This so-called “natural horsemanship” encourages owners to take a spiritual voyage with their horse to attain a new and more humane level of enlightened relationship. As a psychologist with a specialty in learning theory and equine psychology and behaviour, I can assure you that there is nothing magical, elusive, or even “natural,” about natural horsemanship. Most of these trainers rely on well-researched learning principles of classical and operant conditioning introduced by Ivan Pavlov (remember the salivating dogs?) in the early 1900s and taken further by the psychologist B.F. Skinner in the early 1960s.

Natural horsemanship methods are not necessarily harmful when applied with the timing, experience and skill of an experienced trainer. The marketing spin that sells these techniques as revolutionary, magical and embedded in a horse’s inherent desire to please, however, is misleading. Here, I explore the methods of three of the greats of this movement (Monty Roberts, Pat Parelli and John Lyons) and explain what the research tells us about their claims. But first, I’ll sketch in what the last 100 years of psychological research has revealed about how most animals, including humans, learn.


Natural horsemanship trainers use basic learning principles similar to those used to train dogs to sit, animals to do fantastic tricks for movies, children to be civilized citizens and people to give up food, alcohol and cigarette addictions. Their primary learning tool is operant conditioning, which consists of reinforcement (both positive and negative) and punishment (both positive and negative). Almost all horse training (including natural horsemanship) is accomplished through negative reinforcement.

When considering positive and negative reinforcement, think of positive as the addition of something and negative as the removal of something, rather than in value laden terms of good and bad. (Note that negative reinforcement is not the same as punishment).

Positive reinforcement refers to the addition of something pleasant (such as food) that rewards the desired response, making it more likely that the behaviour will reoccur. Our horse bangs his hoof against the stall door at feeding time, we feed him and we make it more likely that the horse will bang his door in the future. Don’t be fooled that since you don’t reward your horse every time, the behaviour will eventually disappear. An intermittent and random reinforcement schedule makes it even more likely that the behaviour will persist in the future (consider our response to the randomly rewarding slot machine!).

Patting your horse and giving him a treat on his way out of the ring, though undoubtedly welcome, is not positive reinforcement. It is unlikely the horse makes an association with a particularly good jumping effort and the reward he gets at the end of the class, making that good jump more likely to reoccur. The food reward would have to be given at the peak of the horse’s jump in order for the horse to understand the specific behaviour that is being rewarded. (There are ways to do this, but I will save that for another time). Positive reinforcement is actually used rarely in most traditional or natural horsemanship methods.

Negative reinforcement refers to the removal of something unpleasant to reward the desired response, making the behaviour more likely to reoccur. We apply pressure with our legs to the horse’s sides, the horse moves forward and we release the pressure. The horse learns that the more readily he moves forward to the leg, the more quickly the pressure will be released.

Punishment (for our purposes we’ll stick to positive punishment) is the addition of something unpleasant in order to make the unwanted behaviour less likely to reoccur. A horse stops at a fence and is hit with a stick in an effort to discourage this behaviour from reoccurring. Punishment is also used in traditional and natural methods, both intentionally and inadvertently (such as when a desired response is not rewarded with a release).


Most natural horsemanship trainers begin by chasing a horse in a round pen, until the horse responds with particular signals, and eventually willingly follows the trainer. Roberts claims that this process of “join-up” occurs because the horse comes to respect the trainer as his dominant leader. Parelli also claims that his methods will teach you “the secrets of becoming your horse’s trusted alpha,” by learning to communicate with horses as they do with each other (as noted on his website). The scientific literature on equine psychology, however, suggests that it is unlikely horses consider the trainer as another dominant horse. Nor is there evidence that natural horsemanship communication mimics the way horses communicate with one another in their natural world.

When round pen training has been conducted under scientifically controlled conditions, researchers found that horses go through a predictable sequence that parallels Roberts’ claims: turning an ear toward the trainer, reducing the diametre of the circle, licking, chewing and stretching, and eventually turning in toward the trainer.

Konstanze Krueger tested 26 horses using natural horsemanship techniques (allowing the horse to investigate the round pen, then chasing the horse in a circle until licking, chewing and stretching was observed, where trainers ceased chasing, and when the horse turned in, offering the horse the option to follow). Krueger found that the time till following (ttf ) decreased daily over 10 subsequent trials, until horses followed the trainer immediately without chasing. Interestingly, ttf decreased predictably whether or not the trainer used vocalization, or a lead rope to chase the horse, with or without wither scratches and regardless of which trainer chased the horse. Although, all of these variations have been purported by Roberts to be critical elements of successful “join-up,” these horses joined-up regardless of how they were chased, or by whom. Also noteworthy, is that when these same “joined-up” horses were given the option to follow the trainer in a small paddock or larger field with or without other horses present, not one of the 26 horses did so (Krueger, 2007). If these horses had come to see their trainer as another dominant horse, surely this relationship would extend beyond the parameters of the round pen.


Also questionable is Roberts’ and Parelli’s claim that their systems mimics how horses communicate with one another – the “language of equus” (Robert’s, 1999). Although anecdotal evidence suggests that horses do chase one another in encounters related to dominance, this is not a consistent finding in research on feral horses, where herd leaders are rarely the most dominant or aggressive.

Natalie Warren-Smith and Paul McGreevy (2008) observed how mature mares coupled with young horses would respond in a round pen on a first encounter. Interestingly, none of the 12 mare/young horse pairs showed the dominance and submission behaviours that Roberts and others have described. The horses licked, chewed, or lowered their heads rarely, and only did so when they were facing away from one another. Furthermore, in situations that might call for submission, such as after the mature mare had made an aggressive overture towards the younger horse, these behaviours were never seen. These researchers suggest that the round pen behaviours horses exhibit when interacting with a trainer might be alternatively explained as a physiological response to a perceived predator. Increased adrenal activity associated with the flight response may lead to a dry mouth, and the licking, chewing and head lowering may indicate an attempt to redistribute saliva in the mouth and nostrils (Warren-Smith & McGreevy, 2008).


Even if we accept the claim that there is some sense of equine social order in the round pen, there is no appropriate analogue to this relationship once we get on a horse’s back. Andrew McLean and Paul McGreevy note that horses mount other horses only in juvenile play and for purposes of copulation, and that it is unlikely that horses interpret our presence on their back as playmate or sexual partner. Furthermore, any ability humans may have to mimic equine communication on the ground, changes once we mount, when the horse can no longer see these same visual cues. Although working with horses in-hand has benefits when the horse is ridden, this may have little to do with dominance. “It may simply be that an unconfused horse in-hand, is a better prospect for training than a confused one” (McGreevy et al., 2009).

John Lyons offers an alternative to Roberts and Parelli by framing his training in straightforward terms of pressure and release. He is thereby able to make a natural transition from the round pen to riding. In his “give to the bit” exercise, for example, introduced initially on the ground, he applies pressure to one side of the bit, when the horse softens or moves his head in the direction of the pressure, the pressure is released. Once the horse learns the exercise on the ground, the same mechanism works when the rider is mounted (from “Your dance partner” on his website).


An alternative explanation to the reconstruction of equine social dynamics, is that horses learn through negative reinforcement how to avoid being chased in the round pen. The horse is able to remove an aversive stimulus (chasing) by directing their attention increasingly toward the trainer. The chasing ceases when the horse turns in and eventually follows the trainer, makes it more likely that turning-in and “joining-up” will reoccur the next time.

Rather than speaking of a “language of equus” or love, loyalty and respect, Lyons explains that any “language is just a series of cues” (Lyons, 1998, p. 27). We can communicate most effectively with our horses, who are more sensitive to body language than words, through our movements and postures. Lyons describes a system of negative reinforcement in the round pen. If we step toward the horse’s nose we tell him we want an outside turn; step to the side and we open a path for the horse to move toward us, facilitating an inside turn. The horse “learns the pattern of your request and, when he gets the right answer, that you leave him alone for a moment” (Lyons, 1998 p. 30). As Lyons says, “the reward is the release … Giving the release at the appropriate time is the key” (From “Your dance partner”).


Perhaps the strongest selling point of natural horsemanship, and the one that seduces so many horse-loving amateurs, is that natural methods are purported to be more humane than traditional methods. Roberts sets up a rhetorical “straw man” argument by positioning his training method against the brutal horse breaking of the wild-west. The “traditional” methods Roberts describes, however, are seldom seen in modern day horse operations, as less aggressive methods are arguably as fast, and certainly less costly in terms of horse and rider casualties. Roberts poses his method against a clearly inferior one that, in fact, no longer exists (1999).

Roberts’ and Parelli’s claims that their techniques do not use force are also questionable. Roberts believes that horses actually appreciate the work he does (as noted on his website), and Parelli states that he avoids using force in favour of “love, language and leadership,” where both parties are “working in an equal partnership” (Parelli’s website).

Round pen work is, however, coercive and based on a horse’s fear response to being chased. Although Parelli and Roberts undoubtedly possess the skill to read a horse’s body cues accurately and instantly, thus making the horse’s stress relatively short-lived, this may not be the case when the technique is used by less experienced amateurs. Furthermore, we use horses for our own pleasure and enjoyment; activities that do not provide the same value for our horses – an animal that has evolved to spend 16 to 20 hours a day on large ranges eating grass. No matter how good our intentions, or how extensive our kindness to our equine counterparts, the fact that we are their “owners” speaks volumes about the “equality” of this relationship. McLean and McGreevy take a particularly hard line in considering the purported kindness of round pen work:

The deliberate use of the flight response in escape-learning contexts by New Age trainers and so-called horse whisperers and their use of … euphemisms to camouflage chasing and consequent fearful responses in the round pen should be eliminated from the enlightened trainer’s tool-kit” (McLean & McGreevy, 2010).

Lyons offers a realistic perspective about the inequitable horse/human relationship. Lyons likens the relationship to a business deal where one partner has the ideas (the human) and the other has the money (the horse). He states that our horses work for us in some capacity that carries no inherent value or logic to them, and it is our job to convince them to be a willing partner in this one-sided agreement. He states, “You would like to do something with the horse that he could really care less about (jumping, racing, cutting, trail riding, etc.).You need to convince the horse that your idea is a great one, one that he wants to be a part of. … A successful partnership depends on you being the most consistent partner and the most forgiving partner” (Lyons, “Your horse is your partner” from his website).


A particularly misleading aspect to much natural horsemanship training is the bestowing of human qualities, characteristics and motivations onto the minds and behaviours of horses – what psychologists refer to as anthropomorphism. This ultimately does a disservice to horses as it creates expectations of human abilities that horses do not possess and overlooks our horses’ true talents and qualities.

For example, Parelli urges would-be followers to pay $189 for a report of their horse’s “Horsenality”, or, for twice the price, a matched “Horesenality/Personality” report for both horse and owner. Although some terminology has been changed, the Horsenality measure is lifted from Wiggins’ (a well-known personality researcher) Circumplex Model of Interpersonal Relationships (2003). Parelli’s claim that knowing whether our horse is right- or leftbrained, or introverted vs. extraverted will be the key to effective horse/ owner communication is truly a stretch. And there is certainly no evidence that Wiggins’ model has ever been validated on an equine sample!

Parelli even claims that cribbing can be cured with his methods and his supplements. He states that if we “love and understand our horses,” feed themhis supplements and train them in his natural horsemanship methods, “your horse’s cribbing problems will naturally disappear” (Parelli’s website). Although there may be anecdotal evidence of cribbing being cured through management modifications, research indicates that cribbing is associated with an endorphin release that becomes self-reinforcing. As such, cribbing is almost never “cured,” even when the environmental factors that precipitated the behaviour have been ameliorated. Nor am I familiar with a literature supporting the notion that cribbing can be terminated through love!


Whether natural horsemanship trainers have managed to recreate, in the human/horse dynamic, an equine social order is questionable. They have, however, successfully capitalized on critical equine social structures that have made horses so amenable to training, namely: 1) the horses’ need to be part of a herd and not be left on their own, making join-up an attractive option; 2) their propensity to form social bonds, preferably to another horse, but to a human if a horse is isolated in a round pen and; 3) that as prey animals their inclination is toward acquiescence rather than dominance. Indeed, it is no accident that horses came to be domesticated and grizzly bears did not.

There is certainly a greater likelihood that a well-mannered horse on the ground will be a well-mannered horse under saddle, and we would not go wrong by introducing more ground-work into our daily routines with our horses. Let us not be fooled though, by the mystique that enshrouds many of these natural horsemanship practices; mysteries that are easily demonstrated in sound empirical research, more parsimoniously and arguably more compellingly.