Ethics

There are three main categories of training within the horse world, each come with their own sets of preaching, justifications, and reasons why they are the “right” method to use with your horse. Various trainers have nuanced each style and try to sell it as their own, using specific tools, names for skills, styles of handling, and keeping of horses – but they all share the same foundations. So let’s look at this analytically, let’s strip away the pretty language, the theories and ideas behind why their techniques work, and the well-proven FALSE dominance theory. So what’s really happening with each of these styles? From a science perspective, how does each teach a behavior?

Traditional training was started when horses were divided out from livestock and began to be used as modes of fast transportation and skilled warfare. This relies on the classic use and understanding of Negative Reinforcement (increasing the frequency of the behavior by removing an aversive). They apply an aversive stimulus directly to the horse, when the horse responds accordingly the aversive is relieved. It’s extremely straight-forward.

An example would be squeezing or tapping legs on the horse’s sides, when the horse moves forward, the squeeze is released. The horse learns to avoid the discomfort by moving forward. This is basic yielding to pressure. The pressure/stimulus used in the training MUST be aversive to the horse, it may be very mild, but it must be something the horse dislikes enough that they are willing to work to avoid it. If the stimulus is not aversive the horse will not work to avoid it, won’t work for the relief of it. This is Relief not Reward, this is utilizing escape/avoidance in training.

Natural Horsemanship is an evolution of traditional horsemanship, with a goal to be kinder and more species appropriate and for the horse as an individual. Unfortunately it is riddled with romanticized misinterpretations of how horses behave in nature. They also rely on the outdated and misunderstood concepts about dominance (about this here: Dominance). They attempt to train in a way similar to how horses communicate with one another. Unfortunately we aren’t horses, horses don’t think we’re horses, we physically can’t take most horse-horse communications, and horses don’t ask anything of each other (like standing tied, riding in circles, or using aversive tools on one another) they only ask the other to “stay away from my resource”. However, this movement has had great aspirations and focus on owners learning to train and work with their own horses. So while much of the foundational information is misguided, the results are forward moving and helping move the horse world towards it’s goal – ethical horsemanship.

So let’s look analytically, how does Natural Horsemanship train behaviors? Ironically, despite all the fancy words, it’s not all that different from traditional. They still apply an aversive stimuli, when the horse responds as desired, the aversive stimuli is relieved. So how is it different? The types of aversive stimuli are different, rather than always applying direct painful pressure (like a whip smack, spur poke, or bit pull, kick…) they may use other options like work (being chased around a round pen a signature of NH) or threats of aversives. These warning signals are another signature of natural horsemanship. This is where they condition a benign signal to predict an aversive, so eventually the handler can use gentle cues instead of always relying on the aversive cue.

This is done by using the non-aversive cue, then the aversive steadily increasing until the horse responds as desired, then the aversive is removed. Soon the time between the warning signal and the strong aversive shrinks, the horse learns to respond quickly to the warning signal, to avoid the aversive stimuli. So while they still use negative reinforcement, they also utilize classical conditioning to train the horse to respond to a gentler cue so we don’t need to use as many actual aversives. However, unfortunately we’ve learned the emotional reaction in the brain/mind is still the same, whether the stimuli is aversive or just conditioned to predict an aversive.

So really, in the thousands of years of working with and training domestic horses training has changed shockingly little. Even the tools have barely changed. We took nose rings and put them in their mouth instead, to make for easier steering from their back… But that was a few thousand years ago. We still use whips, bits, spurs, heels, hands, ropes, and “work” as aversive control devices for our horses. Whether we give them fair warning and use aversives in a wide variety of ways, it’s all the same basic principle. Negative Reinforcement.

So then what is Positive Reinforcement and how is it different? First let’s remember “positive” and “negative” are “adding” and “removing” not “good” and “bad”. Negative reinforcement is removing something the horse dislikes (an aversive) and Positive reinforcement is adding something the horse does like (an appetitive). So positive reinforcement training techniques involve feeding or otherwise giving the horse something they want, when they do the desired behavior. This means we first need to find a way to get the horse to do the behavior we want, so we can positively reinforce it. We have a few techniques for this, capturing (waiting for it to happen and catching it), shaping (reinforcing small steps towards the end goal), and targeting/luring (following a target or the food to guide them into the goal behavior), these options are limited only by your creativity and how well you know your horse. This new approach to working with horses has flipped the horse world on it’s head. Everything is now backwards, horses seeking instead of avoiding, horses rushing TO the arena, hoping training never ends, getting too excited to play with their favorite humans!

While R+ is new as a horse training method, it’s actually not all that new. These learning quadrants have always existed, even before we understood and labeled them. But marine mammal and exotic animal trainers have been utilizing R+ as training tools for decades. Using Negative Reinforcement limited exotic animal training to only what you could use to physically control the animals, which is difficult with large predators like tigers and marine mammals like whales. While possible, it’s impractical, tricky, and very dangerous. Positive reinforcement allows trainers to teach animals without needing to have physical contact or confrontation with the animals they’re working with. In fact they can teach from the side of the pool or the other side of a fence. Even some dog owners are now using remote control video camera treat dispensers to reinforce their dogs for being good even when their person isn’t home! Dog owners were the next to transition, while there’s still some use of aversives, most domestic pet owners utilize positive reinforcement for their training. Not just your classic dogs and cats being trained with treats, but also all sorts of brilliant, exotic birds, rodents, rabbits, bugs and even fish! Now if a wild, dangerous hippo can be trained to hold their mouth open for dental work, a shark to station in a basket for medicine, a lion to offer their paw for blood draws, giraffes to hold their feet up for trimming…. Why on earth would we be resistant to using this kind and forward thinking approach with horses?

Punishment is reinforcing to the punisher. The girls were talking today about their friends who ride traditionally, even those that know about R+, but they use whips and harsh bits and even spurs if their trainers tell them to. Its hard for the girls growing up here to understand why their friends, who are good people and do really love their horses, would knowingly hurt their horse. Because using aversives, punishing the horse, is reinforcing to the rider. We kick, the horse goes, this is fun for us. The horse does something we don’t like, we swish a whip at this and scold them, the horse stops doing the thing we dislike. This is reinforcing to us. Controlling an object or animal through force works – it gets us what we want, and often very quickly, this reinforces the person in control. So while the love and care may be there, it doesn’t outweigh the reinforcement history of having fun riding and being in control.

But what happens when a behavior that was reinforced stops being reinforced? What happens when the horse finally has enough? When they say “no more”?! Well the same thing we see in any animal, the behavior has an extinction burst. We call this a “temper tantrum” in laymen’s terms. When something that used to work stops working. We simply do it bigger and more extremely than before.

So when we are reinforced by using aversive control methods, punishment or R-, when it stops working, we escalate it. At some point it likely will stop working, because animals aren’t inanimate objects, eventually they get sick of getting pushed around. While they may remain compliant for a while, things change. If a pain issue pops up, if a strong distraction happens in the environment, if we finally ask for too much… the horse may finally say “NO”. At that point, the extinction burst happens.

Even for the olympians. Even for the people at the top of the horse industry, they are subject to the same emotions inside all living beings, the same behavioral science. They too have extinction bursts, temper tantrums, when a behavior that was reinforced stops working. So they get out their switch and beat the horse harder.

How can we expect to teach our children empathy, compassion, and doing the right thing (even if the wrong thing feels better)? When the top role models of the industry demonstrate these values? Aren’t we supposed to remind children to make kind and healthy choices, even if the wrong choice sounds more fun or cooler (drugs?)?

Traditional horsemanship relies a great deal on punishing unwanted behaviors from the horse. It can be hard to shake these habits even if we’ve switched to more positively reinforcing training methods. Sometimes when we are confronted with a behavior we dislike we instinctively want to punish them, especially if the behavior is potentially dangerous or makes us feel afraid. Being afraid often triggers our feelings of needing to defend ourselves through fighting back. But we have some major pitfalls that come with using positive punishment that could actually be far more dangerous than finding another way to communicate that we dislike that.

First it’s important to remember that horse’s aren’t being “good” or “bad”, they are simply responding to their environment with their best guess as to how to get what they want and avoid what they don’t. If by performing a behavior we dislike (bucking) finds them relief from what they don’t like (the rider falls), that behavior is being reinforced. The horse chose the “correct” behavior to fix their problem, even though we might dislike it. So it’s important to remember that behaviors are happening for a reason other than “he’s a jerk” or “she’s just being fresh”, horse’s aren’t inherently bad, they are just solving their situation. So the behaviors we are trying to reduce are behaviors we “dislike” or are “unwanted”.

But if the behavior is “working” for the horse, earning them escape from something they dislike (R-) or getting something they want (R+) the behavior is being reinforced! So instead of any sort of punishment we need to make sure to reassess whatever might be reinforcing the unwanted behavior. Look at the environmental situation and find what is reinforcing the behavior. Remember not to just look externally, but also internally, they may have a physical discomfort that this behavior is resolving or an emotional reason this feels good. Maybe being locked in a stall has them pent up so they kick their door? Or maybe they have a stomach ache and kicking their walls is a good frustration release. If kicking their door gets them fed first, guess who’s always going to kick their door? Waiting instead for them to stop before feeding, or removing the door and using a stall guard, or turning the horse out so there is no door to kick, or feeding before they start kicking… These would all be ideal ways to stop reinforcing that unwanted behavior.

We can’t always wait out an unwanted behavior allowing it to extinguish without reinforcement. Some behaviors are even self-reinforcing, they just feel good to the learner. In these times we can use our redirect or reteach methods. If we can redirect our horse to an appropriate behavior and reinforce that soon they will learn to go directly to that more preferred behavior. Such as teaching a horse to station at a target during feeding time. Focus on what we do want our horses to learn instead of just getting mad at what we dislike. Punishment only tells them what not to do, it doesn’t give them an option that we prefer, so even if they stop pawing they may chose another unwanted behavior to fill the void (head tossing, wall biting, bar grinding…)

Remember positive training doesn’t mean permissive training, we don’t just sit there and allow our horse to bite us, drag us, kick us, or otherwise cause us harm, waiting to click when they stop. Instead we Reassess the problem, Remove the reinforcer, Redirect the learner, and Reteach a preferred behavior.

With Positive Reinforcement we finally have the ability to tell our horse “YES”. We can tell our horse exactly what behaviors we like from them and give them incentive to do it. This makes it very easy to mark when they make good choices and communicate clearly what we want. This conditions all our training as something appetitive. As the only stimuli we’re including is appetitive. If they aren’t doing the right behavior, we simply ignore, redirect, or adjust the antecedents, we simply use the lack of “yes” to mean “no”…

Traditional/NH training gave us a clear “No”, we apply aversive sensations until the horse responds as desired, then we remove the aversive. So in this case we say “no” until they guess right then the lack of “no” means “yes”. This conditions our training as aversive, because the only piece we’re adding in the aversive, even though we also remove it. The experience felt unpleasant, even though they found escape in the end.

This seems like two halves of a whole right? Why not use both? We would think this would give our horses more clarity if we can use “yes” and “no”, right? Rather than two halves, they are actually opposites, they work against each other, not together. Behaviorally they both encourage the learner to do the desired behavior and not do the unwanted behaviors, but emotionally we see fallout. The aversive value and the appetitive value of the stimuli we add contradict one another, they don’t add clarity, they add conflict. Because one situation feels good and the other feels bad, putting them together, simply feels confusing, uncomfortable, and conflicting. In simple terms, a threat + a reward may get quick behavioral response, but not a good emotional response.

When we add both “yes” and “no” to the equation, both an appetitive and an aversive, we will create an emotional contradiction within the animal. We’ll see this in the form of conflict and avoidance behaviors, latency before the appropriate behavioral response, and reluctance to accept some reinforcement. If you start seeing conflict behaviors, avoidance, them not wanting the treats, being latency responding to your cues, or any disfunction, look for where the aversive might be sliding into your training. Maybe it wasn’t an intentional aversive, maybe they have a cracked tooth (like one of our ponies!) or something else going on. But if you’re tempted to help your horse out by adding a “no” into the equation, instead look into ways to set the scene to help make “yes” easier to find. Arrange your antecedents, use your shaping tools, and improve your timing, rate and criteria.