Ognen Duzlevski bio photo

Ognen Duzlevski

Senior tinkerer.

Email

(This article is motivated by my training as a computer scientist who has always had an interest in cognitive science. As a horseman, I am always looking for ways to do things better with my animals).

It occurred to me the other day that a lot of what we do as people is habit. We get up in the morning, have our morning coffee, we have it a certain way or we go out for a run or maybe read the newspaper with the breakfast. These are all examples of comfort-driven habits that we have built over time to suit us. There are also habits that get built as responses to pressure and situations we find ourselves in, where comfort is not the goal, but rather survival drives the reactions we produce. For example, some people lash out when threatened, some get their heart rate up and fear response kicks in and for some, they go catatonic when faced with pressure. These reactions tend to turn into habits very quickly and after a while it is not easy to discern with any amount of certainty whether our personality drives the responses or perhaps our responses shape our personalities. In other words, if we react fearfully to one situation once and then we react fearfully to it again and then again, and then a new situation presents itself - do we tend to react fearfully again? Based on personal observation, I would say that these things are highly contextual and that a simple determination cannot be made easily. For example, we may react fearfully when riding a horse and the horse gets upset and we feel like there is no control but we may react aggressively and boldly if someone comes up to us and demands our wallet. Hence, we can perhaps conclude that, at some level, habits form in contexts and are driven by many factors, some unknown (personality?) and some easy to pinpoint - amount of education related to context, amount of training related to context, number of similar situations we have been through etc.

To put this more in context (no pun intended!), we may have trained in a martial art for twenty years and responding to pressure of someone demanding our wallet may elicit a violent response on our part. However, we may be very poor at math and being put on the spot in a math class in school may turn the said black belt into a fearful, low-confidence individual. A few of the above wallet situations may create a very strong and confident person able to survive in any neighborhood and a few of the poor experience math situations above may turn that same person into a very low confidence, poorly performing student.

What all these situations have in common are initial responses and conclusions drawn from them. We tend to learn best from experience and if experience has shown us that a certain strategy works, we will tend to use it over and over. Conversely, if experience has shown us that we have failed multiple times in a certain situation, we will tend to avoid it; we may also develop coping strategies to mask our failure.

I believe that the same thing happens to horses in training (and just in their daily lives). It is often said that horses exist in the moment, they do not scheme and plan and they do not have agendas. This is certainly true, however, it does not negate the fact that horses learn from situations and that they learn to recognize contexts and hence, they will tend to reinforce their behaviors resulting from these contexts, just like people do, especially if left to their own devices.

In practical terms, what does this mean? A horse may be “looky”, “spooky” or “hot” when outside of an arena, for example. To me, a looky horse is a horse that has learned to cope with external stressors in a certain context by becoming overly sensitive to these stressors. If left alone, he will tend to compound the “looky” behavior and perhaps become “spooky” or even “hot”. To get a bit fancy and scientific, horses tend to be simple (or sophisticated, depending on your outlook) neural networks that are constantly pattern matching on context and input. Once there is a match, they act automatically, based on what they know and what has been taught/trained to them. This is what it ACTUALLY means when people say “horses live in the moment”. It is also what it means to train them - we are constantly adjusting their neural networks to respond to contexts in certain ways. This is also why good horsemanship has always emphasized consistency - you cannot “train” a neural network to produce the same outputs for the same inputs, by always changing things - in living beings, the brain’s neural networks by definition look for patterns they can recognize, in order to recognize them (this is a very important thing!). This paragraph has enormous implications on how we train horses but also on how we see them and how we recognize their problems.

Now, onto a practical example - a horse that we have been working with (that had all sorts of issues), who is a very calm animal in the arena but gets very “looky” as soon as he passes through the gate, in order to go on a trail ride.

I will ignore the chemical aspects of a horse’s life here (does he get enough exercise, does he get to be turned out with other horses, is he high on sugar and grains or does he get his nutrition by grazing etc.) - all of these will to some extent modify the magnitude of horse’s reactions but will not change the nature of said reactions. In other words, a horse that spends six days of the week in a stall, being fed grain and doing nothing, will be “amped up” with energy and “high” on sugar, hence his reactions may be heightened in any direction, he may “over-react”, run faster, do more, but at the end it will be the same underlying reaction (fear, boldness etc.), just amplified by the amount of energy that needs to be burned off before a horse can pay attention to his handler.

Back to the story above of the horse whose behavior changes once he gets through the gate. The first thing to ask is what the gate actually signifies and why the horse can be calm one foot inside the arena and “looky” one foot outside of it. These, indeed, are the basic, most important questions to ask when re-training this animal. (Unfortunately, in my experience, most people ignore this change in the animal and ride out on a “looky horse”, which further reinforces to the horse that he is, indeed, the “looky horse”).

In this case, the inside of the arena is the place where the horse’s handler has felt much more comfortable about training the animal - there is a fence (which makes everyone feel safe), there is even ground that is not slippery (which makes us feel confident about not injuring the horse or ourselves if things get ugly), there is a sense of purpose and a good selection of exercises that we can do with the horse to make them better (assuming we know how to do that). Hence, the handler gains confidence, the world is “smaller” in the arena, things are easier to control, but things are also “repeatable”. If the horse “acts up”, we can always dismount and run the horse in circles (for those who believe this is a training tool) and do things from the safety of the ground. So, conceptually, the first difference between the outside world and the arena is NOT the gate, it is what the gate represents - the loss of control, the loss of good footing, the loss of ample space to train the horse, ultimately, the loss of confidence. The gate also represents a bigger world in terms of stimuli for the horse - if you think about it, for most horses training starts in a round pen (small, controlled area with less stimulus), then the world gets a bit bigger in the arena and finally it gets to be very big on the trail.

What really exacerbates the problem is that the horse is then treated differently outside of the arena. Now that said horse is “looky”, the handler has a few choices: the first one is to ignore the behavior and chalk it up to the horse’s personality - the horse is then left to fend for himself and try to survive the best he can while trying to get along with the handler and do what he is asked to do; the second choice the handler has is to “give the horse something to do”. This itself is not a bad approach (comes from “idle mind is the devil’s shop” belief) but it turns out it is very difficult to give someone something to do if that someone believes they are in the fight for their very survival - horses who are in that state (and many of them are, all the time, while handled by people - they just don’t show it to the inexperienced eye) will not listen to their handler and will indeed get worse; in this situation “giving them something to do” easily looks like “punishment” to the worried animal. There are other choices that a handler has, one is to dismount and go back to the barn, concluding that they have a horse that can only be used in the arena, for example. In that same situation above, the horse (not the handler) has a few choices as well: 1) get even more worried and become “spooky” or “hot” and 2) fight back - some horses do not take kindly to being put in tough spots. Some horses build on the “looky” to “spooky” transition very slowly and then they blow up seemingly at nothing - this is just years of learning to “cope”. There are a few more choices the horses have, such as “go catatonic” or become “barn sour” or “buddy sour” or (…) but the first two are the most common.

So, how does one solve this puzzle? Well, one first must understand what is happening to the horse in this situation. At moment x, the horse is in the arena, doing what he is asked to do, fairly happy, fairly calm and perhaps even lazy. As the handler turns towards the gate, the adrenaline in the horse starts to rise, the brain starts looking for neural networks that match contexts they recognize. As you are walking to the gate, the possibility of “getting out of arena” context seems to take precedence, however, the horse also has contexts where he was asked to go towards the gate but the handler then decided to canter for a few laps instead of leaving the arena; hence, the horse is still not prioritizing the “getting out of the arena” context, however, his gate is getting faster, ears are up and pointing forward and he is getting difficult to get a response out of, quickly and softly. This, ladies and gentlemen, is the first sign of what you are going to be dealing with, outside.

There is a good exercise, I believe, every horse should know and every handler should have in their repertoire: it is called “give me your ear”. As every good horseman knows, a horse with an ear on the handler is a horse that is in tune, listening and waiting for a command. Due to the nature of their brain, horses cannot pay attention to two things at once - even when they appear to do so, they actually do not. For example, everyone has had that one horse who could be “hot on a trail”, looking at something intently with ears forward, be asked to side-pass and could do it at the same time as he was staring at the same thing with his ears forward. This does not mean the horse is a multitasker who can pay attention to two things at once - all it means is that the horse is NOT paying attention to the handler and is STILL paying attention to whatever he was staring at anyways.

So, “give me your ear” is an indication of whether your horse is paying attention to you or to something else. If you so much as lift your hand or lightly touch their neck or move your leg, your ideal horse should flick his ear back to you. Even if it is a temporary, two-second kind of a thing, it means that you got their attention. It also means you CAN GET their attention. Incidentally, you will find that getting their ear is much easier in the calm world of the arena and much more difficult in the crazy world of the outside of the arena. This is where the crux of your control lies - a horse who will not even consider giving you their ear will never consider giving you anything else.

Hence, the first step in training your horse (after you have gotten through the mechanics of turns, flexion, departures in different gates, whoa etc.) is the building of various contexts that have one thing in common - “give me your ear”. The first of these is on the ground, then in the round pen, then in the arena. You cannot get a good turn if your horse will not give you their ear, then drop their head and then bend in the rib cage. If you get all three, you got a well trained animal who has gone through the progression of “you have my attention boss, I am dropping my head ‘cause I am relaxing and I am bending in my rib cage ‘cause the whole situation is making me feel so good”. Training for trust and safety, then, is teaching as many contexts as you can where the horse responds the same way. You will find that if he goes through the progression of the three steps above, you can start using turns (and side-passes) as tools that relax the horse by getting his ear and his attention. You have to build this context in the arena, at the gate and outside of the gate. You will find that many horses can do this well inside the arena but poorly just a few steps outside of the arena. Do NOT ride the horse on a trail that cannot do this well outside of the arena, you are asking for broken bones and bad horses and it is all your fault!

Finally, how does a neural network get trained in a living being (such as a horse or a human, for that matter)? Consistent repetition of successful steps - meaning, consistent presentation of the same stimulus, consistently waiting for the horse to relax and only then continuing to something else. What does this look like when retraining the “excited through the gate horse”? Well, it means teaching the “give me your ear” and “bend to relax” exercises where the horse is comfortable (arena?). Then it means getting to the gate and staying at the gate until you can get the same responses at the gate. After that, leaving a few steps through the gate and insisting on the same exercises a few steps out - as soon as the horse gives you what you want (ear, bend), you go back to safety (arena) - this is the “approach and retreat” step of the training, where we consistently build confidence in incremental, small and successful exposures that do not break the horse but habituate him to feeling comfortable and understanding he will not die (this builds trust in us as handlers as well). You will see that actually cracking the nut of the “first twenty steps through the gate and out” is the biggest obstacle. Once the horse creates successful and calming responses to these contexts (and each step out is a context, at least the first 15-20), you will find that now all the “outside of the arena” contexts look the same to the horse. In other words, his brain is finite and he cannot store millions of different contexts, each different by a step he took outside the gate. Hence, after storing 15-20 of these, his brain will generalize into one context for outside of the arena and one for inside the arena. If you can get the same responses outside and inside (same softness, same feel, same mental relaxation state), the horse will then “blend” the context of “inside” and “outside” of the arena into one - “living well in the world” context. This is when rushing through the gate and “lookyiness” will stop.

Stay tuned for explaining how to teach “give me your ear” and how to teach the “bend and drop your head” exercises.

A few corollaries to the above discussion:

1) To point out the obvious for the folks who may have missed it: horses act the way they do out of habit. Hence, “Looky”, “Spooky”, “Hot”, “Lazy”, “Explosive” etc. are all horses who have habituated themselves (or have been allowed to habituate through our ignorance or lack of ability to notice things or even lack of care - some people like “hot” horses, for example) to survive in certain contexts in a certain way. The good news is that, “Great”, “Bombproof”, “Willing” and “Amazing” are also habits…

2) Another corollary to the above article is that “trust” (of horse in handler) is the mathematical set of all contexts where the horse was presented with a situation and survived. The more of these we build in our horse, the better he will be for us.

3) The final point I want to make is that many times it is actually better to take the horse straight from the round pen into an open space and forego the arena (you can and should go back to it). This way, you get to teach all the important stuff in one big “world” context. The arena, then, becomes a specialization of said context for the horse, he will see the arena as a small version of what he has seen and survived “outside”.