Adrienne Schmitke and dressage horse, positive reinforcement

Trained Hound or Treat Hound? Evaluating Positive Reinforcement Training

Science in the Stable

In my life, I have had a lot of exposure to traditional training methods, some of which shun the use of treats for training purposes in horse riding. Rewarding behaviour just was not a thing we did. Maybe they got a carrot after their ride, but that was the extent of the discussion.

Ignoring the fact that positive reinforcement does not necessarily mean treats, I wanted to dissect this notion and lay out what the science says, and intertwine this with the traditional methods we see so commonly in our horse stables.

Adrienne Schmitke and her young dressage mare

Adrienne Schmitke and her dressage mare.

A quick backstory on me before we jump into science. I have trained with Olympians in all 3 olympic disciplines. I have been very lucky to have received a high level education in horses throughout my life. I also believe in evidence-based practice, both as it relates to my profession [manual therapy] and animals. 

Once upon a time, I did not believe positive reinforcement training had any place in my high level jumper program.

My horse did not get treats all the time. I wanted my horse to want to work with me [this is a sentiment we will touch on again soon]. I didn’t want to have to buy their cooperation. Because what if one time I didn’t have a treat on me? If you haven’t got a nickel for the carousel ride, it isn’t starting, and that is exactly what I equated positive reinforcement training to.

Fast forward several years, and I got an exciting challenge thrown my way in the form of a Caucasian Ovcharka, a rare, ancient breed of guard dog from Russia. These dogs are unique in that their history of over 1000 years dictated that they do not rely on humans for any guidance. They are guarding their sheep on an uninhabited mountain in Russia or Georgia, fighting off wolves, bears and other predators. Can you see why this may create some unique challenges when forced to be raised as a pet in 2023?

Nothing in my life would have been more influential to guide me to alternative methods than having a dog who will mature at 100+lb and acting like a wild bear in my house! The short version of this story, is while these intelligent dogs have a reputation of being ‘untrainable’ and not food motivated, devoting our time to building a relationship through positive reinforcement training [that IS the only way with these dogs!!] allowed us to build such a bond of trust that suddenly this untrainable dog was getting her Advanced Trick title [the first in the world to do so!].

So what does this have to do with horses?

Well,  my results with training this dog encouraged me to challenge my own beliefs on positive reinforcement's place in the stable. I wanted to dive further into the scientific literature available on positive reinforcement. I took a positive reinforcement training course with the International Association of Animal Behaviour Consultants [IAABC], as well as read through some available studies.


What is positive reinforcement?

Operant conditioning is a learning process that discusses reinforcement and punishment to garner behaviours, divided into positive and negative. This does not relate to good or bad, but rather the addition or subtraction of something to get the wanted behaviour.

Positive: Add

Negative: Remove

Reinforcement: Encourage the behaviour

Punishment: discourage the behaviour

four quadrants of operant conditioning, FedUpFred
Four quadrants of Operant Conditioning, as illustrated by FedUpFred

Therefore Positive Reinforcement adds something to encourage the behaviour. Negative reinforcement removes something to encourage the behaviour. In the case of horses, negative reinforcement is very commonly used in the form of pressure. Apply pressure, and once the horse gives, it gets a release. 

You have removed the pressure to encourage the behaviour [bending, for example].

In the case of positive reinforcement, you add something to encourage. For example, when your horse halts, you give him a wither scratch.

You have added something pleasurable [scratch] to encourage the behaviour [halting, by example].

That’s all Positive Reinforcement is?

Indeed. There are many rewards to offer for positive reinforcement. It is not just treats. Though treats go a long way too! Because horses are grazing animals and constantly on the hunt for forage, it makes sense that food is a key motivator. But it is not the only motivator.

Other reinforcements could be: pats, scratches, horse friends, etc.

Clicker training is a popular positive reinforcement tool, and that is because it can be so powerful when done correctly. The use of a clicker is to improve feedback time which helps a horse understand exactly what is being rewarded.

First you get your horse to associate the clicker with a positive reinforcement [often a treat]. Click and treat immediately. Click and treat, click and treat, etc. until there comes a time you can click and the horse looks for a treat. Voila. This starts as classical conditioning, like Pavlov’s dogs, and becomes operant conditioning when the horse starts to intentionally repeat an action to illicit reward.

Example:
Your horse holds his head very high and is hard to put the halter on.

Traditional pathways: you may throw the leadrope high up on the neck and try to physically pull to lower the head [negative reinforcement: you’re removing pressure to encourage the behaviour]. You may yell at him to lower his head [positive punishment: you’ve added a vocal response to get him to stop the behaviour]. Or you may leave him in the paddock and take his friend first instead [negative punishment: you’ve removed a friend hoping that will get him to stop the behaviour]. 

Positive reinforcement pathway: you stand there with the halter. Any time your horse drops his head, even if it’s to try to look around you or sniff the ground, or whathaveyou, give him a click, and reward. The first time might take your horse completely by surprise. Stand there and the moment he drops his head at all, click and reward. Soon it becomes engaging for your horse, as he tries to find exactly what causes that bountiful click to occur. Drops his head again and BOOM, another click and reward.

Once you start, you will see just how smart and quick horses are to pick these up! 

horse ears

This is a good time to mention that positive reinforcement training requires learning on both parts. It is a training skill, and a person needs to take the initiative to learn, accept that it is not the horse that is not doing it correctly, but the person. Clicker training helps with timing. Timing is everything with positive reinforcement training. Though that statement could be broadened to say that timing is everything with regards to training. A slow trainer leads to a confused animal, regardless of what method you’re using. Horses put A and B together, a good trainer clearly lays out what A and B are. A slow trainer will have the horse guessing what A and B are. 

 ‘I want my horse to want to work with me’

Remember this statement from earlier? I told you we would come back to it. If you feel that rewarding a horse for the behaviours you want is somehow cheapening the work they do for you, please read on.

Do you remember when you worked for free? Maybe. Maybe there was a time you went above and beyond, or overtime, or did extra. But more than likely that is not what you do everytime you show up for work.

Why should your horse work for free?

The best way to have your horse want to work with you, is to make sure something is in it for them. A lack of something is not the same. A lack of pressure, a lack of fear or pain, is nice, but not equal to a ‘paycheck’.

A study done by Innes and McBride (2008)[1] evaluated feral ponies trained with positive reinforcement and negative reinforcement over a 7-week period. The results ‘suggested that animals trained under a positive reinforcement schedule were more motivated to participate in the training sessions and exhibited more exploratory or ‘trial and error’ type behaviours in novel situations/environments’.

 Trained Hound or Treat Hound?

There is a common misconception that if you have done positive reinforcement training and do not have a treat, the horse will not work with you. That is simply not true. Positive reinforcement helps you build a foundation of trust with your animal such that they feel safe to do as they are asked and are likely to repeat the behaviour even if it is not rewarded each time.

Another study, conducted by Larssen and Roth (2022)[2], found that positive reinforcement training increased the horses’ contact-seeking behaviours.

If you are interested in building a working relationship with your horse, having your horse want to work with you, you may consider adding positive reinforcement into your training. I highly recommend taking a course with a clicker trainer in effort to make sure you are not giving mixed reward signals. Either horse or dog, the principles are the same. Learn the timing and the principles are translational to any animal you work with in the future!

Find a local trainer near you, or consider an online course. There are several offered by IAABC for the public.


Happy training!



[1] Innes, L., & McBride, S. (2008). Negative versus positive reinforcement: An evaluation of training strategies for rehabilitated horses. Applied Animal Behaviour Science, 112(3-4), 357–368. https://doi.org/10.1016/j.applanim.2007.08.011 


[2] Larssen, R., & Roth, L. S. V. (2022). Regular positive reinforcement training increases contact-seeking behaviour in horses. Applied Animal Behaviour Science, 252, 105651. https://doi.org/10.1016/j.applanim.2022.105651 

Back to blog

Leave a comment

Please note, comments need to be approved before they are published.