Pressure….and RELIEF

A little while ago I read an article from a Natural Horsemanship instructor in which she wrote a description of how this method of horse training works.

She said:

“Essentially, Natural Horsemanship methods work on understanding the principles of pressure and release and how to carefully time the application of pressure and the release (or reward) to enable the most efficient learning with the least amount of confusion to the horse.”

This apparently simple description has some omissions and a common flaw. Like many people who want you to follow their way of training horses, the description given of aversive stimulation with negative reinforcement (which is the correct way to describe pressure and release) is that the removal or reduction of the aversive (the release part) is a reward to the horse. It emphatically is not and we have the work of many neuroscientists to show that now without any shadow of doubt.

Natural horsemanship methods universally involve three uses of aversives. They don’t have a monopoly on these though of course – many other traditional and classical and straightness training methods, and techniques promoted by followers of Equitation Science, do the exact same things.

The first is aversive stimulation to produce desired behaviour, with negative reinforcement of the correct response. An aversive (“pressure”) is applied to produce some approximation (a “try”) of desired behaviour and any behaviour by the horse that is in the right direction of the desired behaviour is reinforced by the removal or reduction (the release) in the salience (the strength as perceived by the horse) of the aversive.

This is experienced in the brain of the animal as relief.

The process for this is that the animal experiences the onset of something unpleasant applied by the trainer – something that motivates the horse to take action to escape. As soon as he acts in the way the trainer wants, the horse will experience the removal or reduction in strength of that “something” unpleasant. The feeling that goes with successful escape or avoidance is relief.

The animal then, after a few repetitions, becomes vigilant, hoping to notice any warning sign that the trainer might be about to apply the aversive again, so that this time he can act to avoid it. This warning – whatever the trainer does before applying the aversive – becomes the command to the horse that tells him that he can avoid the actual aversive by acting when he see or feels the warning. It’s sometimes referred to as horses learning “what happens before what happens happens” and the process by which the warning command is learned is a form of classical conditioning called fear conditioning. The horse learns to fear the signal that precedes the aversive onset and acts as if it were the “real thing”.

The second way in which an aversive (pressure) is applied in these aversive forms of horse training, is in an attempt to discourage unwanted behaviour. The trainer tries to ensure that the onset of an aversive stimulus is experienced by the horse as an immediate consequence of any unwanted behaviour. That kind of treatment would be described as positive punishment, if the tendency of the horse to perform the behaviour in that situation decreases.

I was with someone recently who had, like many, been encouraged to hold out her arm to “block” a horse that was going to bite while she was saddling, so that the “horse will run into your arm”. That would be considered positive punishment if the tendency of the horse to bite in that context reduced or stopped, having experienced an aversive (pain) from coming into contact with the long bones of a person’s arm as an immediate consequence of him attempting or threatening to bite.

What this technique didn’t do was to address the reason why the horse felt the need to bite in the first place and of course while the horse might suppress his desire to bite, when the threat of that arm being held out to block him exists, it can result in a whole lot of other avoidance behaviour or displaced behaviour or redirected aggression instead. Or the horse might just shut down – go into a state of learned helplessness, once he learns that nothing he does results in escape or avoidance of aversives.

This is because this blocking tactic does not address the underlying issue – it doesn’t address the reasons why the horse was motivated to bite in the first place.

Similarly “bumping” a horse with the rein or half halting, or hitting (or threatening to hit) the horse on the side of the neck or head with a stick, or using the rein attached to the bit to make the horse go back on the rail or on track if he comes off track when riding in an arena would be considered positive punishment, if the behaviour of coming off the rail or off track (on a circle for instance) reduced in frequency.

If, when he stays on the rail we leave the horse alone, he of course experiences no aversive consequence (other than that experienced from the effort of trotting around the rail, should he have sore feet or legs or be unfit or have an ill fitting saddle or bumpy rider) but neither does he experience a reward. If he comes away from the rail, and we make it “uncomfortable” or threaten to do so, he experiences the onset – the addition – of an aversive as an immediate consequence, and that can act to discourage the unwanted behaviour. This is positive punishment.

So, these aversives-based methods use a mix of consequences.

They use aversive stimulation to motivate the horse to perform a desired escape or avoidance behaviour, with negative reinforcement (stimulus removal or reduction) to make that behaviour more likely to be repeated. And they use the immediate onset of some kind of aversive stimulus – attempted positive punishment – to try to reduce or eliminate unwanted behaviour. I use the word “attempted” intentionally because often the behaviour is just suppressed only when the potential for that aversive to come on exists, and actually, because the motivation to do it remains, the behaviour is rarely actually eliminated.

The other way in which aversives are applied, is in an effort to get a horse to accept stimuli or situations it finds aversive and to cease responding to them. We want the horse to accept and ultimately ignore irrelevant things. That could be traffic, other animals, noises, general commotion, tack, rugs, yard tools or equipment and so on.

Sometimes the horse is exposed, at liberty, unrestrained and in an open area where he is free to choose to leave, to a potentially aversive stimulus – a novel object or potentially frightening situation at very low strength, and then perhaps the stimulus is removed – or the horse is removed from the situation – when the horse relaxes. The stimulus strength is increased very slowly and gradually, all the time making sure the horse relaxes completely before proceeding. If that’s done well it can be an excellent way to get horses used to new things without them experiencing fear or feeling trapped, or giving up trying to leave because their escape attempts are futile. The horse gets to choose to leave if it’s too much for him. Ideally, if it’s done well he will never choose to move away, never to return.

More commonly however, the horse is restrained on a halter and lead rope or in a bridle and confined in a small pen or in a stable or trapped up against a wall and subjected to an aversive event or stimulus until he ceases to try to escape, whereupon either the strength of that stimulus is reduced or the stimulus is moved away or the horse is let out. This is called flooding, with negative reinforcement, the horse being removed from the aversive stimulus or situation – (or vice versa) to reinforce a reduced response.

Sometimes called approach and retreat, other times called overshadowing, this is in almost all cases flooding with negative reinforcement. With a horse physically restrained or believing himself to be so, we should always assume that the horse will experience this as flooding. Flooding is routinely used in an effort to have horses habituate (have them learn to tolerate and cease to try to escape) to being weaned or removed from the herd, to being stabled, tied, travelled, saddled, being exposed to tarpaulins or balls or flapping flags or coats or clippers or water or veterinary procedures.

It is not very often that you will see a horse being exposed to aversives for habituation purposes in a wide open space where they could leave should they find the situation too much.

I’ve recently been asked to help someone who has a horse that is fearful of being stabled – it could be fear of separation from the herd or fear of being confined or the horse could have had some past experience of aversive treatment in the stable or over the door. In this case it’s probably a mix.

It could even be likened to PTSD (post traumatic stress disorder) as the horse was force-weaned from his mother and that involved locking him in a stable for days on end.

Despite being stabled daily for some years, the horse is still stressed in the stable and tries to get out of the stable whenever the door is opened, whether someone is in the way or not. The horse will follow a person into the stable and he will go into the stable if pressure is applied to send him in there, but he immediately tries to come out.

I always take history of the horse including medical, management and training history. When asked about what she has already tried, the person involved has tried natural horsemanship for quite some time. The approach recommended was to have the horse on a halter and lead rope in the stable and to use a stick to block the path of the horse and to make the horse back up when it tried to leave the stable by hitting him between the front legs on the chest. This was described to the person who owns the horse as pressure-release and that when the horse backed up she should “reward” the horse by ceasing to hit the horse on the chest.

Bear in mind that the horse is afraid to be in the stable and that he is going to experience the addition of an aversive (being hit on the chest) when he tries to escape the stable. When he backs up away from the door that aversive (pressure) will be removed to reinforce his behaviour of backing up. Having followed this advice for a while, now, when the person opens the door to go in with the horse, the horse shoots to the back of the stable and shakes or turns his hind end on her and threatens to kick. She decided she doesn’t want the horse to do either of these things.

For a horse that fears being in the stable and isolated from friends, we have to ask the obvious question, which is this:

In what way will experiencing relief from pressure to back up away from the stable door ever be experienced as a “reward” for that horse, when the reason the horse is doing this is because he is stressed about being isolated in the stable?

Indeed for any horse, whether he is just minding his own business standing quietly in the stable looking out of the door, or whether he wants to come forward to escape what he perceives to be an aversive situation, how could we have ever believed that not being hit on the chest anymore with a stick because you just backed up would be experienced psychologically as a reward?

A person with an education in Psychology and who is up to date with the latest information we have from the science of Affective Neuroscience would be able to explain that the removal of an aversive stimulus (the release of pressure) is not ever experienced psychologically or neurologically as a reward. At best it is relief. Other neuroscientists would describe any use of aversives as punishment regardless of whether such a stimulus were being used to produce or stop behaviour. In any event removal of one type of aversive may be no relief at all, if the horse is still in a situation he finds frightening or uncomfortable or painful anyway.

People who repeat this information – that escape or avoidance of aversives is a reward – do not always intend to mislead you. They are just repeating what they’ve been told by someone else who perhaps has no qualifications in horse behaviour or psychology or neuroscience and who probably hasn’t read about the emotional aspects of learning. The emotions the horse experiences when you stop hitting her with a stick or loosen the rein are not experienced in any way at all like the emotions experienced when she is given a carrot. Not even remotely.

Please share this so that we can educate the novice student of the horse that the escape or avoidance of something unpleasant is never experienced as a reward.

It’s called pressure…. and RELIEF.