Monty Roberts on Positive Reinforcement training

I just finished reading an article by Monty Roberts, where he said that he would not recommend training a horse with treats (ala clicker training). While he liked the idea of a marker (clicker), he was adamantly against the use of food. I have to admit that coming from such a great horseman, this is very disappointing to read. I understand that he had to deal with a clicker trained horse and that it was very aggressive and biting. In fact, he stated that "when food is associated with the human body it produces horses that bite." So in his mind, it was the food based positive reinforcement that produced this problem.

First, I really thought he would know better, after all, he is a master of timing, and that is what this is all about. Treats do not create a biting horse, it is the poor timing of ANY reward or reinforcement that creates a bad behavior. This includes the timing of "release" in modern horsemanship. Secondly, clicker training has only been around a short while, and before that, there were plenty of dangerous horses being created by people attempting natural or traditional horsemanship. Monty has been working with problem horses long before clicker training came about, and not all of them were created by spoiling with food. Food is not the devil for horsemanship. When you understand animal learning theory, you know that food is just a primary reinforcer, and when given during a specific behavior, it *will* reinforce that behavior, because food is highly desired. This is what makes it so powerful with horses. As such, what truly matters is WHEN the food is offered. For example, if a disrespectful horse invades your space and mugs you, and you immediately give him food, then you have rewarded his muggy, disrespectful behavior. Clearly, if you continue to do this, it could escalate out of control. An experienced clicker trainer knows this theory and avoids this situation.

In modern horsemanship, the same theory applies, only in this case, the method is using negative reinforcement. (Negative meaning something is removed, not that the method is "bad"). The rider removes pressure when the horse performs a behavior. The horse then learns to repeat his response to avoid pressure.

In either method, a poorly timed treat or a poorly timed release will get you things you don't want.

Regardless of which style of training a person ends up using, criticizing a training method, unless it is outright abusive and cruel, serves nothing and only limits us in what we can achieve with our horses. Everyone has a different take on what should be taught and how, but that doesn't mean someone else's way is inferior or bad. After all, when you see someone bring in an poorly trained horse and they said they followed the Monty Roberts method....you would be a fool to say that Monty's method was terrible. In reality, someone with poor skills has only their skills to blame, not the style of training. Sure some styles might be easier to follow than others, and a horse and person might excel under one method but not another, but all training is dependent on feel, timing, experience and even the horse's personality and previous training.

Monty's implication that clicker training is not going to fix horses like "his method" isn't an appropriate criticism either. The ladies he invited didn't necessarily work with dangerous horses on a regular basis, therefore it is questionable that they could fix that problem quickly, regardless of what method they used. The process of undoing damage isn't going to be overnight, even for an experienced trainer. And just because you see an instant result with a training method, does not mean that the problem is now gone. Most behavior fixes require long term consistency to really stick. As soon as the horse goes back to the original owners and isn't managed like the expert, the problem will resurface. Therefore, his article's subtle criticism of the quality of the clicker training method seems obviously biased in favor of his own school of thought and does not consider these other factors.

It would have been much more professional for him to say that clicker training, like any method of training, is acceptable in the hands of someone who knows how to train a horse and can avoid the caveats and issues that might come up. After all, any method under the wrong hands can lead to problem behaviors and every method requires more than just basic knowledge to have success.

And whether you use Monty's method or the Clicker method, if you don't know what you are doing, you shouldn't start training a horse to begin with. It is not advisable for anyone to commence training a horse until they truly understand learning theory and have thoroughly researched how their chosen method works and receives expert guidance. That means a lot of books, videos and lessons/clinics, all by experts. Otherwise yes, you can train your horse to be dangerous with any method.