Positive Reinforcement Horse Training from a Dog Trainer’s Perspective

Written by Christina Young, CDBC

Using positive reinforcement to train horses isn’t new, but it doesn’t have the same following as it does in dog training. My mare Shea is now 20 years old and was trained in traditional methods based on punishment and negative reinforcement (P+ and R-), although I did dabble in clicker training to teach her a few tricks.  It honestly never even crossed my mind that I could train her husbandry and riding skills using positive reinforcement (R+). 

As my dog training advanced, the traditional ways of training horses became an obstacle to my riding. I didn’t want to ride that way anymore and I didn’t want to have that kind of relationship with my horse, whom I love deeply. I restarted Shea using R+, stationing, and consent. She has always had issues with being tacked up, including biting or kicking out at me when I’d brush her or particularly when being saddled. In the past, P+ was used in a “don’t you dare” type of training, but guess what? I still had to be careful when brushing or tacking her up or she might bite me. Her new R+ program included a station that she could offer to be on to be groomed and tacked up. She was free to walk away at any time. 

With Shea getting older and needing to slow down, I had to make some choices. Was I going to continue training and riding horses? Was it possible to train a horse without using pain, fear, or threats or by creating learned helplessness? I decided that I did want horses to remain a big part of my life, so last summer I brought home a yearling Arabian stud colt who we named Zephyr.  I even started a podcast on our training journey.

With both Zephyr and Shea, I am taking my time to foster and maintain calm and optimistic emotional states. I am hoping to create safe, sane horses who think first rather than react first, who enjoy spending time with me, and who I enjoy spending time with. 

Similarities and Differences

There are a few key differences between horses and dogs that affect how we train. Some of these differences I anticipated, and some caught me by surprise.  

Speed

Horses are slow and respond well to slow, calm movements. When I feed the horses their morning grain, I bring out two buckets, empty one into Shea’s corner feeder where she is standing and waiting, then wait for Zephyr to position himself by his corner feeder so that I can empty his bucket into his feeder. If I am in a rush and move too quickly or pressure Zephyr to move into place quickly, he is likely to become agitated. If I take my time and allow him the time and space needed, he calmly moves into place.

Dogs eat very quickly and move very quickly, so we can get many repetitions of a skill in a short time. With dogs, in 30 seconds, I might have 10 iterations of a training task and reinforce 10 times. Horses, even though they can be fast movers if startled or excited, take time to eat and move. In 30 seconds, I might only complete one or two repetitions of a skill, delivering just one or two reinforcers. 

Consider training a foot target. If my plan is to mark when the animal steps on the target, then reinforce away from the target, setting up a training loop, this can be done very quickly with a dog by tossing a treat on the ground or into a bucket.  Tossing a treat on the ground for horses doesn’t work well in this training scenario as it could take a very long time for the horse to find it. If a horse steps onto a target and I reinforce in a bucket nearby, it might take  two to five seconds for the horse to move to the bucket, two to five more seconds for them to eat the treat, then  five to 10 more seconds for them to turn around and move back to the foot target. 

Session Length

A 20- to 30-iteration session with a horse can take a lot longer than a similar session training a dog. It isn’t unusual for a session with my horses to last for 15 minutes without a rest, whereas I rarely train a dog for longer than five minutes without a break. 

Adjusting the Training Plan

With horses, there is more time to think between iterations and I can adjust my training plan on the fly to account for the horse’s responses. For instance, if I notice that the foot target is too close to a fence for optimal learning, while the horse is eating the reinforcer, I can consider the issue, plan the adjustment, and reposition the target. With dogs, if the training plan needs an adjustment, I often need to station the dog, scatter treats, or add other time-buying games to give myself time to think and adjust my plan. 

Positioning

Because the sessions with horses are slower paced, I have much more time to think and to ensure I am ready and in position for the next iteration. With dogs, I tend to reinforce in a way that sets the dog up to be in position for the next iteration. With horses, I might do this a little, but usually I move myself into position while they are busy moving toward or chewing the reinforcer. For instance, if my body is acting as a lure to entice the horse to move toward  me, stepping on a foot target as they move, it will be faster for me to put my body in position rather than adjust the horse’s position. 

Environmental Distractions  

Most of us take our dogs out into the world to train skills such as loose leash walking. We likely start inside in a low-distraction environment, but taking dogs into the world is generally a primary goal, even for young puppies who need to socialize. Training our dogs to be comfortable with and to work around distractions means that attention to the handler and recall are  a lot of what we focus on.  Dogs spend so much time inside with us, that when out and about, we are not the novel stimuli in the environment. 

A horse’s world is generally much less complex, and the end goal of most skills except for trail riding or showing is to be performed in a fairly controlled environment that  is often in a fenced paddock or arena. I don’t have to train focus or attention so much, as I am the novel / interesting thing in the horse’s environment. In fact, when training two horses at once, stationing and waiting for their turn became a main focus as they both found training very reinforcing. 

Safety

Food delivery:  Many trainers use multiple marker systems with dogs to indicate how the reinforcer is being delivered. A multiple marker system isn’t necessarily required for horses, but teaching them that food will be delivered on the ground or in a bucket is  important, and not one that comes naturally, at least to my horses. We need to be able to reinforce away from our hands and bodies, because although we try to avoid having an animal frustrated or over-aroused in training, sometimes we fail. A dog that becomes snatchy with food, nipping or jumping at hands when over-aroused is typically manageable, whereas a similar behaviour in a horse can be dangerous. If we help people feel safe when training, they will be more likely to continue exploring R+ training options. If people feel unsafe, they are likely to pursue other training methods, and as I see in online chat groups. may actively speak out against R+.  

Protected contact, where the trainer stands behind a fence, gate, stall guard, etc. while training is also very common in horse training. Protected contact is great because the horse learns about markers and how to access reinforcement without the option of pushing into the handler to try  to take the treats from the treat pouch or hands. We use protected contact with dogs with serious aggression issues, and also with small children learning to interact with dogs, but otherwise most of us are quite comfortable with dogs in our space. Horses are big, powerful animals, and the risk of injury for even experienced handlers is significant. 

I love the idea of starting all novice trainers with protected contact until the horse and trainer have the mechanics of how to cue, mark, and deliver/access reinforcement. Here is a video of me training Zephyr. We are advancing our stationing skills and teaching him about multiple marker systems. I am training behind a fence (protected contact) to help him learn to stay a few feet away from me, even when I have treats and we are in a training session.

Consent

Much of the horse industry has a different definition of the term “choice.” Consider a situation where the handler would like a horse to stand in the centre of a set of cross ties to be saddled. A common technique is to tap or apply pressure to the horse until they stand where desired, then say that the horse is choosing to stand at that spot. However, since the alternative is that the handler makes any other choice uncomfortable and the horse does not have the option of walking away, this is not how I or most dog trainers would define choice. The horse is not choosing to participate, just choosing the least aversive of two pressures. 

However, some horse clicker trainers have a very exciting understanding of choice. They advocate for having a treat ball available to the horse, containing the same treats being used in the session. The horse only consents to training if they leave the treat ball to participate in the exercise. This is a much more generous application of a consent test than most dog trainers implement.

Allowing an animal to opt out of a training session is still relatively new for dog trainers, and there appears to be as much or more resistance to this in the horse world. I hope that as we progress, allowing  animals more choice will become mainstream.

Training

Stationing  

I advocate teaching a horse to station before any other skills.  Stationing is useful for many husbandry tasks, teaches the horse that reinforcement comes away from the human/treat pouch, and it teaches a calm default behaviour. It is also easy to build consent into a training session using the station as a start button. 

When training more complex tasks such as standing for eye drops, I can build momentum  via a higher rate of reinforcement by continuing to reinforce an easy behaviour such as stationing in between the harder tasks. Here my friend Kerri-Lynne uses a station to teach her “hard to handle” horse to stand for eye drops.

Stationing is an excellent default behaviour as it is rarely a wrong choice for a horse to stand still at a station. 

Targeting

Many horse clicker trainers introduce a nose target first or very early in the learning process. The targeting behaviour is an easy one for most people to succeed at, giving the handler and the horse a high rate of reinforcement. However, without a defined useful purpose and combined with new R+ trainers who don’t understand the concept of stimulus control, horses offering to target objects out of the context of a training session can be frustrating for horse and  handler. 

I did introduce nose targeting early, but only in the very specific context of targeting the halter. Zephyr was anxious about the halter near his face, so I did a few sessions with him to create a positive conditioned emotional response (CER+) to the halter being held out in his direction. I was concerned that an aversion to the halter could create a horse that was hard to catch, but with just a few sessions, the anxiety disappeared and he became excited and interested in the halter. 

Conclusion

While I love trick training and the power it has to give animals and trainers new ways to think and communicate, I discourage it as an activity for people new to R+ horses. In my opinion, if we want the horse community to embrace R+ training of horses, we need to show them how it can be simple and effective in addressing common training and husbandry challenges without causing frustration or unwanted behaviour problems.  R+ horse training needs to be simple without the need for novice trainers to understand stimulus control, choosing to train only behaviours that would not pose a risk or be an annoyance if the horse offers them when not prompted. If more horse people start using R+ for basic husbandry skills, it can become one of the “tools” they think about when trying to solve their next behaviour problem. 

With my yearling Zephyr, I plan to first focus on:

  • Catching and haltering
  • Stationing
  • Grooming and tacking up
  • Moving away from pressure
  • Walking over bridges
  • Stepping in puddles

I hope that as more of us show the world what we can do with our horses, more people will shift toward a kinder, more consensual relationship with our equine friends. 


Christina owns Positive Dog, specializing in aggression and reactivity. She creates humane, effective, and realistic training plans that help dogs and their families.  She enjoys off-leash hiking with her dogs and trains and competes in many dog sports including Disc, Nosework, and Tracking. Christina believes that training should be accessible to everyone so she offers a free on-line course and support group for reactive dogs as well as hosting two training podcasts. 

TO CITE: Young, C. (2024). Positive reinforcement horse training from a dog trainer’s perspective. The IAABC Foundation Journal 29, doi: 10.55736/iaabcfj29.5

 

SHARE