Articles

THE PRESSURE PRINCIPLE - The removal of physical pressure or discomfort reinforces (rewards) whatever behaviour precedes the moment of removal (i.e. leg(s) rein(s), spurs, whip-tap, headcollar).

The pressure principle is the most fundamental principle in the training of performance horses.  The bit in the horse's mouth and our legs around the sides of the horse deliver various pressures to the horse's body. Our aim is to shrink those pressures into invisible light aids.  However even with light aids there is still a range of light pressure variations. The performance horse is constantly receiving subtle (or not so subtle) variations in pressures that are supposed to translate to variations in mobility. It is an essential part of training therefore to train the horse how to respond to pressures. 

Pressures are learned by negative reinforcement (pressure removal), and then when converted to lighter aids can be maintained by positive reinforcement (reward). When you think of pressure/release, think of Tom Robert's words where he would pose the question "When you sit on a pin, why do you get off?" Most would answer "Because it hurts" but Tom would correct them and say "No, you get off because it STOPS hurting". That provides a powerfully clear message about how the reins and legs work to produce responses in the horse.

TRAIN PRESSURES THOROUGHLY

One of the most important points to make here is that we must train using pressure/release thoroughly, rather than rely too early on fragile associations. A good example is training the horse to lead correctly. Horses are more than capable of learning to avoid pressures on the head collar by simply learning to copy your footsteps. When you move the horse moves, when you stop it stops. Simple. Except when you decide to lead it somewhere it doesn't want to go, like over a creek or into the float. Then the gates of hell open. Nice horse turns into monster and monster soon learns that this is just the thin end of the wedge. How did the adorable Dr Jekyll transform into the evil Mr Hyde? Because our equine Dr Jekyll learned that the pressure disappeared when he went berserk and the dysfunction of signal/alternative response make him insecure because the signal loses its predictability. That's pretty well how all ‘bad' behaviour is learned. He rears - the pressure on the rein and leg disappears; he bucks - the whole problem (the rider) disappears; he shies - the rider loses balance and control. The truth is whatever behaviour immediately precedes the removal of pressure, the horse learns that it caused it. The more flight response involved, the faster he learns it. The more signal/response dysfunctions, the greater the insecurity and the more likely flight response behaviours are to occur.

Thus we need to place boundaries around the horse's behaviour, and this is through the use of the reins and driving aids of the legs. Using our legs to control horses would have been obvious to our ancestors who first tried to control them. But what about bits?

Horsemen conceived the idea of putting bits in the horse's mouth a long time ago too. Apparently the ancient Numidians controlled horses with only their legs, but it wasn't a successful recipe for control especially in crisis situations. The Institute for Ancient equestrian studies reveals that progress with deceleration control came early: there is indirect evidence of bit wear on horses' teeth found in the Ukraine from 4000 BC; that antler cheek-pieces were used as anchors for rope and that hide or sinew mouthpieces have been recovered at sites on the Black Sea. It is believed that metal bits originated around 1500 BC. It seems we recognised the grim truth quite early on that putting a bit in a horse's mouth made horse riding less of a lottery, safety-wise, than it would otherwise be.

WHY ESTABLISH PRESSURE/RELEASE FIRST?

There are a number of reasons for the priority of establishing pressure/release behaviours first. Perhaps the most important is that correct pressure/release provides solid boundaries for behaviour. This is essential for a ridden animal that can gallop at 70 km/hr, weighs around 500 kgs can kick at the force of 1.8 times its own bodyweight yet has no comprehension of consequences, but rather learns by the somewhat risky means of trial and error!  It is small wonder that in the Western world, the annual death rate is one death per million head of population. The bottom line is you have to get the horse under control. With correctly trained rein responses you have a greater chance of stopping a bolting horse more than any other way. With correctly trained leg or leading responses, you have the greatest chance of making it go across that creek or into that float when it just won't budge.

Secondly the correct use of pressure/release is efficient. It rapidly induces the behaviour that we target as our response. For example we want the horse to step sideways from the tapping of the whip on his hindleg. We increase the speed of the tapping until he moves. If he moves toward the whip rather than away from it we can increase further the speed of the tapping. We can pay ‘colder-warmer: you got it!' with pressure/release and achieve results rapidly, thereby lowering the number of incorrect repetitions. 

Another advantage of pressure/release versus mere associations is that operant learning (trial and error, negative and positive reinforcement) is more stable than unrelated associative cues. Everyone who lunges horses will be aware of the fact that horses rapidly learn to go forward and to slow on voice commands. Yet you still need to have that lungeing whip nearby because every now and then, the voice command fails. The less similar a cue is to the pressure/release effector, the more easily it is forgotten. So the visual picture of a lunge whip tucked under your arm is more effective than voice, but it too is less effective than the whip itself. 

Finally when pressure release is correctly trained, it begins to achieve a reliable pattern of response in about 5 repetitions. That is mighty fast by any measure of learning. Successful trainers of young horses will know exactly what I mean here. When a horse first learns to stop from rein pressure, it takes only about 5 repetitions and the pressure used from then on is almost always light. The difficult part is actually getting the pressures right. There are no established national institutions that teach us how to use pressure/release correctly. It is one of those things that is thought of as an art, a gift so most people blunder along not knowing how to use pressures correctly so they avoid situations where they might need to. The horse soon learns to dictate his own pressures. The horse teaches the rider not to use his reins, leg or whip. ‘Hot' horses are adept at making the rider keep the legs off. Horses are wonderful animals and make great trainers!

NOT ALL HORES ARE THE SAME

Horses vary tremendously in their responses to pressure. This is related to their sensitivity, and just how aversive they find the particular pressure. Some horses need very little increase in pressure, and tend to offer the correct response almost at the first experience. Others need more pressure before they are tempted to respond, and others respond but offer all the wrong responses as they mentally trawl through all their entire possibilities before coming up with the right one. Horses that take a while to offer the correct responses or to offer any response at all are often referred to as being stupid. Yet when you use food as a reward as I have done with some of my experiments on mental abilities in horses, the ones that seem stupid with pressure are frequently Einsteins when it comes to learning that involves food rewards. This shows that what we are really testing when we use pressures is not intelligence but motivation.         

SPEED OF RELEASE

Horses have enormous difficulty in learning something when the reward is too many seconds away from the behaviour. In fact, optimal learning occurs when reinforcement or reward adjoins the correct behaviour. You have to be quick. This is also one of the reasons why giving the horse one good whack is unenlightened. Chances are, this sudden and dramatic increase in pressure that amounts to punishment will result an explosion of the wrong response, rewarded by removal of the whip. Now you have really trained the wrong thing. Some years ago, researchers Haag, Rudman and Houpt demonstrated that punishing horses lowers their ability to offer a new behaviour in solving a problem. Similarly Professor Daniel Mills, an eminent ethologist, pointed out that punishment in training is problematic because it tells the animal what not to do, but not what to do. He showed that punishment has the potential to desensitise an animal to the punishing stimulus if the punishment intensity is not correct. Mills also pointed out the risks of deleterious emotional changes that can interfere with attention and learning, and the fact that punishment may be associated by the animal with the person delivering it. He concludes that punishment is a dangerous minefield of problems that amount to abuse, and are best avoided.  

MOTIVATING PRESSURE

Careful use of pressure-release however is an entirely different thing. If you think of pressure on an arbitrary scale of 0 to 10, you will better understand the correct use of negative reinforcement. Level 0 is no contact, i.e. a loopy rein, or legs away. Level 1 is contact (neutral). The horse has to habituate to this ‘neutral' level of pressure. Level 2, 3 & 4 are the various doses of the light aids that eventually elicit all responses. Levels 5 to 10 are the increases in pressure and these are intolerable for extended periods. In training and re-training, the way pressure should be used for all responses, both rein and leg, is that it should reasonably quickly escalate from the light aid through to the stronger pressures then release the instant the horse gives the correct response. The release needs to be abundantly clear to the horse, however after a number of successful repetitions the release doesn't need to be a loss of contact, simply reverting to the contact level (1) is sufficient. When you discover which particular level of pressure results in the correct response, you increase to that level each time and SKIP the pressures that don't work until the horse offers the same quality of response from the light aid. So it is critical to always begin with the light aid. This is called the ‘motivating pressure'. So your level of pressure might go 1, 2, 6, 0, 1. This translates as Contact/light aid/motivating pressure/release/contact.

TIMING

For effective habit formation, the training of transitions (therefore the use of pressure-release) should be within an identical time-frame. Without such a time-frame, transitions take far longer to acquire. Therefore optimal training also means that the pressures should be synchronous with the rhythm of the footfalls of the horse. The most effective way to achieve this is to focus on the separate beats of the forelegs and implement the ‘three beats of the rhythm' principle. So a walk to halt transition goes RF, LF then RF again. This corresponds exactly with Light aid, stronger pressure, release.  Importantly, this is the way horses actually learn to respond from light aids, rather than always needing stronger pressure. It is the close proximity of the light aid to the release of pressure that enables all horses to maximally perceive and respond to the light aid. This is as much to do with duration as anything else so in the canter and gallop, the three beats of the rhythm is best applied where a complete stride is seen as one beat - therefore in those gaits the whole sequence is compete in three strides.  Using pressure release this way thoroughly during training obviates the need for stronger bits in cross-country. The event horse that is ‘strong in the mouth' cannot perceive the light aid as it is too far away in time from the release.

Collaborative research with Amanda Warren-Smith and Dr Paul McGreevy (Uni of Sydney) showed that initial acquisition of lead forward responses in foals was most effectively trained after releasing pressure on completion of just one step forward. However once this has occurred, the most effective protocol for longer term retention was to subsequently release on the third leg step which coincides with the format in the above paragraph. This doesn't mean the horse can't learn transitions in any other sequence, it just means that this is the most effective way for optimal learning. And of course it stands to reason that reducing ambiguity has got to lead to better welfare than a hit and miss training strategy that can encourage conflict behaviours.

CONFLICT

Conflict behaviours are those that we commonly call resistances, evasions and bad and dangerous behaviour. These are induced by confusions that arise through poor timing of release or no release at all. Therefore the way in which trainers and riders use pressure determines the horse's future to a very large extent. When you get the pressure/release mechanism right, you provide very solid foundations in training; when you get the pressures wrong the horse becomes confused and can ritualise (make a habit of) his conflict behaviours. That's why good trainers always aim for consistent results each time. They are fussy, and with good reason, otherwise you are rewarding different results each time and the horse doesn't have a clue what you want.  That is also why a good foundation in hand and under saddle is critical: during stress the animal will, like his owner, tend to default to his first habits. 

Pressure itself is aversive or unpleasant to the horse - his aim is to remove the pressure. The more unpleasant the pressure, the more neurotic it can make the horse if the animal cannot find a way, or rather find a behaviour that makes the pressure go away. Pressure without release such as strong unrelenting contacts and nagging legs result in desensitisation of the horse to pressures. Some individual horses can apparently live with it (but not without some negative effect), others cannot. Conflict behaviours such as bucking, shying and leaping, are best treated by deleting them with downward transitions, and then repeating the aid that caused the incorrect response. This sequence is repeated until the correct behaviour is shown and it should be given ample practise. Incorrect behaviours should not be given more than three strides of expression otherwise the horse takes this to be the correct response simply because when the behaviour exceeds three strides it is incorporated into his repertoire.   

Prevention is always better than cure. This is particularly true when it comes to anything to do with the flight response. Once exhibited they are more deeply learned with less practice that the ‘correct' things we try to train horses to do. The proper and humane use of pressure-release is the best tool we have to prevent such behaviours. It is important to train speed and line control in our horses in hand and under saddle with clear pressure-release principles. When these shrink to light cues, we can maintain them through the use of positive reinforcement. We have to be careful not to fall into the trap of trying to control these things with the seat or weight before the horse has learned the boundaries of pressures, and their conversion to light versions of the rein and leg. Then later in training if the subtleties of seat and posture go awry or transitions fall out of the three beat sequence (which they are prone to) we have a back-up plan that works immediately, limiting the amount of incorrect and confusing experience.