The Relationships between Horse Training and Ethology
The equine community is all very aware of horses behaviour in terms of being a flight animal that requires space, forage and company in order to feel safe. But what does natural horse behaviour mean in terms of how we train them?
A common component of many ‘natural horsemanship’ (NH) systems is an emphasis on the necessity of the trainer becoming the horse’s leader by asserting their dominance status over the horse, demonstrating that they can ‘move his feet’ and maintain their own personal space at anytime. This is often achieved through round pen training and groundwork exercises requiring the horse to move in all directions. Although I am very pleased that the popularity of these training programs has resulted in many people understanding about the importance of the release of pressure at the correct time, consistency, how our body’s affect the horse and that good communication does take time and effort: I am more then a little concerned about some ethological inaccuracies in these techniques. Many are considered as humane and natural, but ethological and psychological evidence suggests that this is not always the case.
Round pen training takes place in an enclosure which prevents escape, and is advocated on the basis that it mimics equine social interactions. Horses in round pen training are believed to respond to the trainer as they would to another horse.The resultant partnership, is one in which the horse is believed to be a willing participant, despite its formation being based on driving a horse away until the trainer feels he is ready to submit: essentially stress. This is a vast contrast from the way in which horses form and maintain bonds; through mutual grooming, grazing and play.
The horse is released into the round pen and is first chased away from the trainer by the application of agonistic or predatory postural and auditory stimuli which elicit flight responses in the horse: supposedly asserting the trainer the role of the herd leader or alpha mare by banishing the horse from the protection of the ‘herd’. Agonistic or predatory stimuli may include sudden movements by the trainer, arm flapping, an ‘aggressive posture’, a direct gaze, throwing a rope or line at the horse and loud verbal sounds and are believed to mimic the behaviour of horses engaged in aggressive agonistic interactions. The trainer may force the horse to change direction, demonstrating her ability to move the horses feet. After a period which may be up to fifeteen minutes the intensity of the aversive cues are reduced which enables the horse (by now probably exhausted) to trial additional responses such as lowering its head to the ground, moving in closer to the trainer and licking and chewing, thus signalling respect and its desire to return to the vicinity of the trainer or safety of the herd. The trainer will soften their posture and movements and avert their gaze, mimicing avoidance agonistic behaviour between horses. Some trainers may then approach the horse, stroking it on the nose and then walk away to test the horses’s acceptance of their superior role. If the horse does not follow another chase sequence will begin. This process will be repeated until the horse learns that following is the best option. And there I have said it…the horse learns.
So how do horses decide who is their leader?
Many ethological studies have been carried out around the world and factors which affect rank in a herd include; age, longevity of band membership, maternal rank, social experience and individual temperament. Mares and stallions both were witnessed to hold dominant positions. Positions of leadership may also vary depending on the activity and individual motivations and priorities. Linear dominance hierarchies were often observed but triangular ones were also witnessed. This gives us an idea of the complexity of equine social behaviour and ‘moving the others feet’ alone has not been recognised as hugely relevant in ethological studies of free ranging and wild horses.
I am not a horse. I will not behave like one, and my horse will never see me as one. And that is ok.
I do not believe that dominance or hierarchy are relevant to training as a pre-requisite or justification for how we interact with horses:
The majority of studies of free ranging and domestic horses suggest that dominance hierarchies are stable and rarely challenged. So why would we need to keep enforcing our superiority? A good leader shouldn’t. However, repetition is necessary for good communication, lightness and responsiveness to cues to be achieved. This is important for safety, welfare and fitness. It is also important if we want horses to be able to carry out the jobs we, as humans have assigned them. Therefore we have a responsibility to ensure they enjoy that training and activity and not do it for the sake of enforcing a leadership role which the horse doesn’t really want anyway.
Aggressive agonistic behaviour is significantly higher in domestic groups due to domestic groups being of unnatural composition, subject to unnatural change and exacerbated by an unnatural lack of resources and an inability to avoid agonistic interactions due to spatial constraints. This lack of agonistic behaviour in natural settings brings to question the necessity to use agonistic posturing to gain social advantage over horses during training. Some natural horsemanship and traditional techniques puts an over emphasis on agonistic behaviour which does not appear to be an accurate parallel to natural horse behaviour.
There is a huge amount of misinterpretation in round pen training type techniques. Horses are generally a very passive and co-operative species, and ethological studies show that agonistic behaviour only represents a small proportion of interactions which tend to be limited to competition over resources. The most reliable predictor of the dominance relationship within dyads is avoidance behaviour rather than the expression of aggression by a dominant animal.
Ethological studies have shown that the role of leader will vary depending on the activity. Dominance is often a flexible concept dependent on confidence and needs, and will change between horses, situations and as motivational needs change. E.g. a hungry horse may move another horse away from a pile of hay, but a satiated horse may leave it to munch happily. The social structure has not changed, but the motivational needs have.
Age, sex, length of residence, temperament, size and maternal rank are all factors in leadership roles in free range and wild horses: none of these things can be factors in the human:horse relationship.
So what does the horses’ response to round pen training, and similar ground exercises used to assert leadership, really mean for the horse?
It has been hypothesised by Ladewig (2007) that avoidance of an aggressive horse can be explained as an example of negative reinforcement i.e. a horse avoids aversive conflict by keeping a certain distance, thus avoiding the pressure. Interestingly, some natural horse trainers view avoidance of the trainer as a sign of disrespect for the leadership of the trainer and will be punished by repeating the chasing sequence until they follow.
Round pen trainers will use chasing, direction of gaze, body angle in relation to the horse and speed of approach as postural cues, mimicking natural equine behaviour….
However, chasing between horses would naturally be short duration and cease once the desired result is achieved, i.e. resource is now available or preferred distance of personal space achieved. Training technique set aside, we ask our horses to maintain movement and paces for longer than they would naturally offer, except in situations where they must maintain speed due to predation. Otherwise, they may use their physical abilities through play or to achieve a specific goal, e.g. A to B, or see off another horse. Take lunging, round pen training and Parelli games for example: are very different from moving another equid away from a resource and thus I do not believe it should be considered comparable.
The concept of gaze is subject to debate as it is very hard to gaze at a horse in any particular way absent of any other posturing. Suffice to say, directed gaze accompanied with other agonistic movements is likely to be viewed as predatory or threatening behaviour. Monty Roberts (1997) uses gaze and head height to communicate either aggressive or affiliative intent. Parelli (1993) talks about energy and visual focus on the ‘zone’ that you are seeking to move. This is incredibly effective (I know because I may use focus to move a horse), suggesting that horses are extremely sensitive to gaze and subtle posturing. This is one thing I really like about some NH programs: although I do not advocate pressure being the primary motivator to achieve results, I appreciate that some of these programs have laid out a clear guideline regarding how to correctly and fairly apply pressure, only escalating when necessary and as much as necessary. Studies have shown that horses can discriminate on the basis of a handlers gaze and head position but this fact was proven based on positive reinforcement training within experimental settings rather than response based studies suggesting any ethological significance (reviewed in Henshall & McGreevy, 2014). Speed of approach does seem to be significant in influencing a horse to move, whereas studies showed that body posture was less significant (Birke et al, 2011). This possibly suggests the horses response is more likely to be a flight response then a social response.
Licking and chewing:
This is often talked about as a sign of submission in round pen training (after been chased around a pen), or mental processing in Parelli (1993) circles and I have heard it discussed in some massage/therapeutic practitioners in such a way. Snapping in young horses directed towards older horses has been observed and considered part of the equid ethogram by ethologists(McDonnell, 2003) but no licking or chewing is recorded. There are a lot of conflicting opinions in the literature regarding licking and chewing. Personally, I am in agreement with Henshall & McGreevy (2014): this behaviour is often accompanied with lowered head, yawning and rapid blinking and there may also be some physiological initiators relating to the body’s response to stress. This coincides with a lot of the behaviour work I do with dogs: lip licking, yawning, lowering of the head, averting eye contact, blinking: are all well-known ‘calming’ signals in dogs designed to tell the other dog/person to back off as they do not wish to fight. If this is the case then what we are seeing in round pen training is a stressed animal trying his best to avoid conflict.
Summary
Traditional and natural horsemanship trainers have contributed to our understanding today of horse behaviour in so many ways. Natural horse trainers are skilled in observing subtle body language and utilising the flight response with minimal physical force to achieve results, and in many ways this will improve welfare. But what can be said for his or her psychological well being?
Many people that are attracted to natural horsemanship training are so because they want a deep and emotional relationship with their horses, and they are made to feel weak for lacking the necessary leadership for their horse to feel safe with them. I cannot help feeling they are being miss sold in some cases as it is clear that social organisation is clearly more complex then conditioning responses to cues (using positive punishment and negative reinforcement – see How horses learn page) in order to get the horse to follow, back up, changing direction etc. Horses know that we are not horses. That is not to say that we cannot have a deep emotional relationship with a horse, but it should be based on mutual respect, partnership and understanding: not agonistic behaviour. It is my opinion that all the behavioural responses seen can be explained by learning: not social submission.
Training techniques which take into account the cognitive abilities of horses, what motivates them and what are fair expectations are likely to result in improved training outcomes and reduced performance of unwanted flight behaviour.
Many people say that their horses seems calmer after starting natural horsemanship training. Seeing some of the training, I wonder if this is calmer or simply resigned to the fact that they are not entitled to an opinion, even if what we are asking does not make sense to them. This of course cannot be true for all natural horsemanship style trainers. There are some that take a more philosophical approach. There are some that would have helped many horses and owners just through teaching about consistency, the release of pressure at the correct time, a gradual escalation of controlled pressure rather than using full force and responsibility as a horseman that if things do not go to plan it is up to the trainer to consider how to break the learning down so the animal can understand. That said, I would still prefer to use food and scratches to reward correct responses, rather than apply pressure on my partner. Pressure, at whatever degree, is aversive: otherwise it would not motivate.
HOW HORSES LEARN
Horses learn like any other animal. If an action produces a pleasant outcome, they will repeat that action. If an action produces an aversive outcome, they will avoid that action. Many of the animals we share our lives with have become companions, rather than creatures that are relied upon for livelihood as they often historically were. With the modern cultural shift has come an explosion in knowledge on animal learning and emotion. As such, most people now wish to train their dogs using reward based methods and find the idea of choking and bullying them quite unbearable.
So what about horses? There have been many scientific studies on learning theory in horses, many of which have included reward based methods. Yet the equine industry appears to be resistant to change. Essentially, all the dressage riders, show jumpers, cowboys and various natural horsemanship programs….rightly or wrongly, knowingly or unknowingly, are working with the principles of learning.
It is essential for every rider or horse-person to understand how and what their horse is learning. This seems to be a very muddied area in equine circles as euphemisms are frequently used, resulting in a lack of understanding of equine practices and the psychological and physiological effect this may have on the horse.
So lets just look at learning theory in its most basic form:
A questionnaire showed that 79.5% of equine coaches felt that PR was very useful, but only 2.8% could correctly explain the application of positive reinforcement, and only 7.8% could correctly explain the use of negative reinforcement (Warren-Smith & McGreevy, 2008). This lack of understanding may result in poor training technique including an absence of the release of pressure, release of pressure at the wrong time, opposing pressures simultaneously and the absence of shaping (McLean, 2005), all of which could result in poor welfare and impede learning (Hothersall & Casey, 2011; Mills, 1998).
How many of you have been in a riding lesson and your instructor has asked you to try some leg yielding up the long side and you find yourself shifting your weight, half halting on the reins but simultaneously ‘putting your leg on’ and your poor horse ends up trying a few things and then giving up and just drifting back to the track despondently, probably thinking ‘I don’t know what I was supposed to do but I hope this was right because I’m sick of all this pressure being applied!’. I certainly remember this from my riding school days.
This might get better with practice because the horse learns through trial and error what works and what doesn’t. Good trainers understand exactly what they want and exactly when to release the pressure to communicate that with the horse. The problem here is that through this learning process the horse will get the wrong answer many times before they get to the correct one and therefore the learning experience is less likely to be enjoyable. When there is an absence of shaping and/or no release of pressure a horse will suffer from confusion, psychological and physical suffering and his welfare is compromised. Some trainers also try to hold a behaviour for too long before the horse has had the pressure: release process applied repetitively to understand what it is you are asking for, resulting in confusion the next time round. This is common in many dressage movements. Another common issue is that aspects of riding such as riding ‘on the bit’ result in a constant pressure if not done with lightness in mind: a desirable goal for most riders.
Shaping is where a behaviour is broken down and trained in small increments, gradually progressing to the final goal and is vital for reliable learning and kind training. For example, dressage movements can be broken down into easily comprehensible chunks, or scary requests such as going over a ditch, in a trailer/horse box or a spooky situation could be broken down in terms of distance and intensity: while the horses confidence is gradually build over time. We can’t get it right the first time, and nor can our horses. The nature of horses and the things we ask of them also means desensitisation and counter-conditioning techniques (aka gradually getting them used to something and making it a pleasant experience so they can form a positive emotional association to it) are a frequent necessity: but unfortunately many people prefer to use pressure to force the horse to comply even when they are not ready.
This is what horse riding generally is: pressure being applied in various ways and the horse has to figure out how to make that pressure go away. The pressure is positive punishment (‘squeeze with your leg’, ‘give him a tap’, ‘cheer him up’, ‘push him on’, ‘squeeze on the rein’, ‘ask’, ‘suggest’) and the release of the pressure at the correct time is negative reinforcement. Traditional and natural horsemanship training programs often rely predominantly on positive punishment and negative reinforcement.
In short: this training says ‘No, that is wrong’ until the horse gets the right answer and the trainer removes the pressure.
Positive reinforcement is a term a lot of you would have heard of, probably from dog training. Your horse does something you like and gets some nuts or a scratch for example. This cannot be possible without some degree of negative punishment – i.e. until your horse does something you like the reward is withheld. Learning is not without stress, and even reward based learning can be aversive if it is not done correctly. Laura believes how an animal is taught a skill or exercise will determine their emotional response towards that behaviour throughout their life, as well as their motivation to willingly carry it out.
In short: this training says ‘Yes, you got it right’ and ignores the wrong answers. This will help them find the right answer much quicker.
Some trainers inaccurately say that the removal of pressure (i.e. the aversive) is positive reinforcement. This is true in the sense that the horse will feel relief. However, this does not make the training or learning enjoyable and something the animal would wish to seek again. A good analogy is: if someone were to hold a gun up at you and instruct you to give them your wallet, you would do it, they would remove their gun and you would feel relief. Would you like to be mugged again? Probably not!
Through advances in neuroscience, we now know that reward based training uses the ‘seeking’ circuitry within the amagdala: a primitive part of the brain that is heavily involved with emotion. Activation of the seeking circuit is a pleasurable time for animals, as anticipation of reward is shown to be more reinforcing then the reward itself. Nature has designed us this way so we are motivated to find what we need to survive. When Laura starts working with an animal, she can generally tell how that animal was trained and how they felt about it at the time, even if they were trained years ago. This is really apparent with horses who have been taught to lunge using driving and chase type techniques: you don’t have to do much as one small gesture will send the horse out and they will just keep going regardless. Some might say this is successful training, but if the horse is emotionally unhappy then this is not the case. Teaching based on avoidance responses will only result in one thing: a horse who wants to get away.
An investigation of the behavioural response towards the task and trainer when teaching rein back to equines (Equus caballas) using positive or negative reinforcement
https://www.academia.edu/8632468/Laura_Palmer_dissertation_final