The photo above shows one of the approximations we use when teaching trailer loading. Navigating a narrow space alongside the trailer, confidently, is a major step along the way to gaining confidence to enter a trailer.

Successive approximations, simply put, mean that we start with what the horse can offer already and gradually direct and reward each tiny move in the direction of the final behavior we want.

In other words, at the beginning of teaching something new, we click&treat for the slightest approximation of what we want as our final result. Each improved approximation is one tiny slice of the overall task.

When the horse feels ready, we encourage him to do a tiny bit more to gain the click&treat. This whole process of rewarding successive approximations is called ‘shaping a behavior’.

Putting a behaviour on signal/cue: In the next clip, you will see me re-setting the task repeatedly as I show Boots that a click&treat will only be forthcoming if I give the signal/cue for this behaviour, which is dropping my head and shoulders forward. We don’t want the horse forever putting his head down.

A human example of shaping a behavior is teaching a child to write. The child starts with holding a pencil and using it to make random marks on paper. At some point the random marks become conscious curves and straight lines.

When the time is right, we introduce writing letters of the alphabet. Eventually the child can group letters to make words. Words are then arranged into meaningful sentences. Some children go on to write coherent paragraphs, essays, stories and books.

If the child loses confidence with any of the slices of the process, an element of discomfort creeps in, along with typical avoidance behavior. Not enough practice then results in a poorly shaped skill.

When a horse loses confidence in his ability to understand what we want him to do to earn a click&treat, we get typical avoidance behavior. He shuts down, stops trying – moves away if he can. Keeping the learner’s confidence is the essential role of a teacher.

Writing is an interesting human endeavor that starts at two years old and is still being shaped many years later at high school, tertiary education and beyond. There is always room for improvement.

Another way to look at successive approximations is to think of a sculptor starting with a piece of stone. He works in careful stages until the shape in his mind is visible to the rest of us in the shape of the stone.

In the same way, we gradually tease a series of movements (or stillness) out of a horse to yield the task we want. This is a bit harder than shaping stone because horses have minds of their own!

General Key Points

1. The more quickly you click&treat, and simultaneously stop any signal/cue, when the horse complies, the faster the horse can learn to recognize just what it is you want.
2. Once the horse understands your intent, refine your signals/cues and be TOTALLY CONSISTENT with them.   
3. If you keep nagging with a signal/cue by repeating it, you will desensitize the horse and make him disregard the signal. Nagging is an easy habit to get into and hard to break. If the horse does not understand the first time you use the signal/cue, back up in your teaching until he does.
4. The energy you use to communicate with a signal/cue will change with the situation and what you are asking the horse to do.
5. To teach with the free-shaping, the horse learns without a direct signal/cue from you to initiate an action because you have set up the environment to make the action you want likely to happen. The horse does the action of his own free will. You mark the action with a click and reward it with a treat. Once the horse offers the behavior reliably, add a signal/cue to it. The signal/cue is usually determined by the nature of the task. It will be unique to you and a particular horse. For the haltering task in the video clip at the end of the blog, I started with the horse’s natural curiosity about the hoop, and shaped the haltering procedure from that.
6. Just when everything is feeling really good is the time to STOP. Avoid at all costs the urge to do it again to see if we can. Change to something relaxing instead.
7. One of the key skills of horsemanship is to read the horse accurately to be able to decide on the best activity (or rest) for the moment we are in. It is part of the concept of ‘feel’.
8. When things go awry (which they will), always go back to where both you and the horse feel confident again, and gradually work forward from there.

When we watch a horse and his trainer perform a complex task, it is not always easy to see the steps the handler took to reach the smooth end behavior. Even if we can find out how the horse was trained, that only tells us about what worked for that horse and that handler.

Here are some questions we have to ask ourselves.

• What did the horse know already before learning that task?
• How experienced is the handler?
• What method(s) of reinforcement does the handler use?
• How long have the horse and handler worked together?
• How long did it take them to reach this point?
• What were the tricky bits along the way?
• How often did the horse get confused?
• How often did the handler have to go back a few steps and work forward again?

Seeing a horse carrying out a task that grabs our fancy is the beginning of our own adventure. It gives us the inspiration to teach our horse something new. The starting point for any Individual Education Program we set up for a horse will be unique to us.

Here is the definition of thin-slicing again.

Thin-slicing is a shorthand way of saying: split the overall task we want the horse to learn into its smallest teachable parts and teach each part in a way that makes it as easy as possible for the horse to understand.

Resetting a Task

When we teach something new, we are experimenting with our signals/cues and the horse is also experimenting to work out what it is we want him to do. It’s not unusual for things to get a bit complicated and messy.

If either you or the horse lose track of what you are doing, pretend it was perfect, relax (but no click&treat), pause. Count to ten and rotate your shoulders. Then go back to the beginning of the task and try again, starting with your visualization of how a good effort will look and feel.

The magic about pretending it went well (when it actually turned into a mess) is that it dissolves the natural frustration we feel when our communication is not getting through. Horses instantly pick up on the changes in our body when we are frustrated or angry, and will want to move away from the negativity, so any teaching/learning moment is lost.

If we can smile, breathe out and relax our body before the reset, we don’t upset the horse or make him anxious. We simply start again.

When someone tries to teach me something new on a computer or cell phone, they invariably go so fast that I have no idea what they did or what I should be doing. I think that is what often happens with horses.

The following two clips demonstrate teaching with successive approximations. Our sessions were never much longer than three minutes.

In the next video clip, be aware that to make it shorter for easier watching, I cut out the bits where I reach into my pocket, after the click, so it is a bit jumpy. Everyone’s progress will be different.