“But he KNOWS what I want.” Oh REALLY?

Have you ever heard someone say “But my horse KNOWS what I want!” I now think that common statement can’t be further from the truth. This changed for me dramatically when I started clicker training, and with it, my understanding of how a horse perceives a cue.

So what changed my thinking? The effects of the way I train behaviors with the clicker. Let me describe the process. First, much of the teaching involves allowing the horse to offer behavior, for which he is reinforced.  I do a lot of shaping, which means I teach an animal to guess at what I want and click/treat him for all of the little guesses that are within close range of the goal. Before long I can build a complete behavior from those guesses, including complex riding behaviors.

When trained using this type of shaping, the animal usually loves doing the behavior because there is a strong reinforcement history that is a part of the fiber of the behavior, it is built in. Proof of this is when you even begin to see the animal FREELY offer this behavior in hopes of getting the reinforcement it wants.

The next step of the process is to add a signal to the behavior so that the horse knows when YOU want the behavior to happen. This process of teaching a cue is a bit more complicated and can be a source of confusion for horses and people alike.

Because humans have the ability to process complex combinations of input, we tend to like to give directions that include multiple levels of input. We speak, gesture, nod, touch and more, sometimes all at once. To prove the point, just ask someone to give you driving directions to a place that’s familiar to them.

Because of this human tendency, we often gives multiple signals to our horses, and this can cause confusion. To illustrate this with humans, when you are given directions, which of their signals do you focus on the most? Verbal? Gestures? Visual pictures? How can we be sure which of our cues was significant to the horse?

Another point worth mentioning is that our brain has developed around millennia of complex verbal communications. The horse may understand sounds, but not as we typically use them. Teaching verbal cues that rhyme or sound similar can be a source of confusion.

Aspects that significantly impact our ability to teach our horses to reliably respond to the specific signals that mean “DO this behavior now,” are also ruled by the context in which we teach them. We often teach our behaviors in specific situations and they haven’t been generalized. Horses need to be taught to generalize behaviors as well as their cues.

Imagine standing by a horse and offering the verbal cue “Walk-on” expecting that the horse will step forward. What if the horse takes two steps and stops to touch a target he saw in your hand? Is he being stubborn or disrespectful? Does he know better and is just ignoring you?

Let’s imagine that you taught your horse to automatically go to a place in your barn to stand or be ground tied. He understood that this station was a place where he had gone time and time again and by doing so had earned a lot of reinforcement. Now imagine that as your horse walked towards the station you said “Whoa” before he got there? How would your horse respond?

WHICH cue do you think would be stronger, the visual cue of the station or the verbal cue to “whoa?” Remember, the horse WANTS to earn reinforcement so it is very motivated to pick the right answer.

Back in the days when I used pressure to maintain behavior, a little flick of the rope would have been the follow up to a missed cue. Once I took pressure out of the toolkit as a backup for my cues, I saw how cues can be missed by horses who really WANT to perform a behavior and just didn’t understand the cue, or it was trumped by something else.

The other cue related phenomenon arising in clicker training horses is the situation where people are getting interested in the process, they’re training lots of fun behaviors, they love watching the horse become enthusiastic, but they didn’t follow through and learn the process of teaching cues for those behaviors.

If the horse learned to touch a target, chances are the owners didn’t teach the horse to wait for a SIGNAL to touch the target, and if the owner used a word along with the visual presentation of the target, they assumed the horse really understood the word without testing the cue. The most significant indicator to perform the behavior was the site of the target itself.

So what happens to the horse who is cued to do something else, but sees a target? If the horse isn’t reinforced for that offering, what next? Most likely the horse begins to TOUCH anything and everything in sight in order to get the human to interact and play the game.

This kind of frantic activity startles and even scares some horse owners. They begin to think the horse has a respect problem or they think the attitude is all wrong. They see enthusiasm and sometimes mislabel it as over-excitement, when what we’re REALLY seeing is that natural phenomenon that emerges when an animal is trying to get what it wants, but doesn’t have a clue about the cue. In other words, they work harder at finding a way to get the reinforcement we’ve trained them to expect. I feel a sense of sadness for the horses and how often they’ve been corrected for something that was not clearly taught or proofed.

So what’s the cure? LEARN THE NEXT STEP of clicker training, LEARN how to teach cues correctly. Your horse has an ability to offer behavior, it is a gift, it is precious. Don’t PUNISH that beautiful gift by offering less reinforcement or randomly paying the click to curb his enthusiasm. Learn how to teach your horse the cues that go along with the behavior. Learn how to do this with a few behaviors without using pressure as the motivator. Become proficient at proofing your cues in other places, using good techniques like offering the cue ONCE instead of repeating it, cluttering it up with multiple signals given simulatneously, or having one cue (like pointing) mean five different behaviors.

Once your horse understands the cue, you can change when and how you reinforce it. But please, get educated on this incredibly important part of clicker training.

Don’t try to guess your way through this, speaking from experience, I wasted precious time trying to learn it on my own. Seek the professionals that can show you how to teach cues to clicker trained animals.

Peggy Hogan – the best “whisper” is a “click”

Posted in Uncategorized | 8 Comments

Expanding my understanding of cues

 

Back in 2007 I decided to take on a big challenge and develop a freestyle dance routine with 2 miniature horses, to be performed live at a local horse show.  The process challenged my understanding of cues, cue control (stimulus control), behavior chains and performance at liberty. From time to time I would like to share some training strategies learned from the process with the hopes that you can apply some of the concepts to your own training. I hope you all see that the concepts are universal and apply to all horses. Do not be misled into thinking “Well that only works with minis.”

During the actual live performance, the horses were not completely at liberty in that I placed is very small one foot high barrier of a single white PVC pipes stretched between low uprights; the horses could have jumped them at any point during the routine. This factor added some security, but the behaviors and the reinforcement history of those behaviors was the glue that held the routine together.

McKee was only 3 years old at the time, Handsome was 4, and you’ll remember that they both had behavioral issues when they came to me.

The project was a big stretch for me and it really brought up some interesting training challenges. For example, in the routine McKee had several behaviors that involved doing something with his mouth. He had to bite the corner of a blanket and pull it off of me, he had to bite a handle of a coffee cup and throw it in a trash can, and he also had to retrieve a small feed tub from the ground and hand to me several times in a row. The tricky part was helping him understand which behavior to do when.

You see, most people, including myself, would think that an animal could easily differentiate between items and tasks, and automatically learn the subtle nuances, but that was a lot to assume on my part.

So during the training, after I shaped each mouth oriented behavior, we went through a phase where McKee offered to do every single behavior with every single item; it looked like a three ringed circus. I consulted Ken Ramirez. The advice included training each behavior in a separate location, at a different time of day., then slowly integrating them into the routine both in time of day and location. Who knew? Simple, yet complex in terms of training strategies.

After we had performed the routine, I filmed it at home. The following link takes you to the video so you can see the behaviors I described. More installments to follow. Enjoy.

Posted in Uncategorized | Tagged , , , | Leave a comment

Is it too much work to carry treats for your horse?

From time to time someone will ask me about the need for continued use of reinforcement in their training program. Basically I think people tire of having to focus on their horse and what motivates the horse. They also might grow weary of carrying treats with them when they train.

My reply involves asking people to make a choice based on knowledge of theory as opposed to cultural horse norms. In the science of training there are two ways to maintain, or “reinforce” a behavior. You are either going to add something that the horse likes or you are going to add an aversive and remove it when the horse complies. I’m talking about the difference between positive and negative reinforcement or in popular circles it might be seen as clicker training compared to pressure/release training.

One of the reasons behind the concern expressed with continuing to offer reinforcers is that people have some underlying misconceptions that behavior should become self-rewarding or the animal “should” enjoy doing his job. People make this assumption for a variety of reasons. One of the most deceptive is in the comparison to pressure techniques. You see, the culturally familiar tool of habitually used pressure tends to make people forget it is still pressure. In other words as people, we’re used to seeing horses trained with pressure, so we just accept and assume that the horse enjoys the work if he’s not resisting.

When the horse begins to be light in his response, we even more easily forget that we are using the quadrant based on aversives or the threat of aversives and so therefore think we are not reinforcing a behavior with our lead, whip, rope or spur. It’s as if the behavior is just taking care of and maintaining itself.

By comparison, when we reach up and give the horse a bite of food, it takes conscious attention; it makes us more aware of the process and we might even perceive it as excessive in comparison to the often unconscious use of pressure when we are reinforcing a behavior that is in place.

With the act of using actual food, we might even be told we’re doing something “less” powerful or effective than with other methods.

If you want to do a little comparison on your own, go take a walk with your horse. Stroll out for 200 or maybe 300 yards and watch how many times your hand touches the lead rope in some sort of maneuver.

Do you use the lead rope to suggest that the horses turn? Do you use lead rope to suggest that he slow down? Even lightly, do you use a lead rope to suggest that the horse give you more space as he walks? Just watch your motions and observe them.

How many times do you touch that line in order to give the horse direction? Even the common “wand as an extension of my arm” mindset can play itself out as lots of subtle gestures that keep the horse in place or moving at the chosen speed.

In my journey with clicker training I approached this leading process very differently. My horses and horses owned by people I teach, have learned to shape (hands off) every aspect of the leading behavior. The horse learns to regulate his pace walking with us, not by pressure, but by being clicked and rewarded for regulating his own pace.

The horse learns to stop, back up, and perform yields of every kind. I also train transitions of walk, trot, and canter. The horses learn, through the use of the positive reinforcer, how to walk next to me comfortably and safely.

As the behavior matures I slowly vary the reinforcers for individual behaviors and I begin to have more behavior sequences. I also continue to train other behaviors that get reinforced more frequently. However, I still reinforce each individual behavior from time to time. As a result I can walk with my horse in a variety of situations with a variety of distractions, never touching the lead rope (usually we’re at liberty) and reaching up to reinforce the whole leading behavior with a bite of food once in a while.

So I would ask you to compare the differences in terms of choosing a reinforcer. On well-learned behaviors, I still reinforce with food. Yes, it takes conscious thought. In comparison I see a continuous succession of prompts with pressure handlers.  I see a lot of subtle rope checks or use of those wands or whips. So I guess it’s a matter of how you want your horse to perceive the tools for strengthening a behavior.

That makes the choice easy for me.

Posted in clicker training horses | Tagged , | 5 Comments

I got no strings on me.

Who needs whips or ropes?

Peggy Hogan – the best “whisper” is a “click”

Posted in Posts for fun | Tagged , | Leave a comment

Horses who “target” too much, and the women who love them.

There is a HUGE attitude change, from the horse’s perspective, when shifting from other training styles to clicker training. One of the main reasons is that when there is something IN IT for the horse, he wants to REPEAT what worked for him. Right? So he’ll offer to do everything he knows in order to get the person to start delivering reinforcement again.This can be a bit disconcerting if you’re not used to seeing a horse offer behavior (unless except when he’s worked up, and then he might be offering behavior you don’t want!). However, when you’re clicker training, you want the horse to offer behavior, especially when you’re in the learning phase of a behavior. After the behavior is learned, you need to know how to teach the horse specific signals or CUES that let him know WHEN you want him to do the behavior.

A great example of this is targeting. A lot of people, including me, teach the horse to target in the beginning. You see, once he learns that targeting can earn him a click/treat, he’ll just keep touching the target, usually pretty quickly, because it worked for him, and the visual presentation of the target is a cue, in and of itself. Make sense? In fact, we could walk away and show him some hay and chances are he’d touch that target a couple more times just to see if it makes the click/treat thing happen again. Why mess with a sure thing!

So if you find yourself in the luxurious spot of realizing your horse is offering a behavior over and over, YOU need to take the next step and learn how to teach him a special signal that means, “Touch this now, and only when I ask.” It does take time, skill and understanding to teach this to both human and horse. This step is a VERY important part of clicker training. There are some frustrated people AND horses out there because the people didn’t know how to support the horse in learning this central point of good clicker training. Don’t be one of them. Learn the ins and outs of good cue control.

Posted in clicker training horses | Tagged , , , | Leave a comment

When different species interact

Sometimes animal interactions just tickle me down to my toes. Today’s giggle came a little tiny bit at the expense of one of my minis, Handsome.

I went to the upper paddock to bring Handsome and McKee down to their more sheltered evening paddock. While I walked up there, I heard a Scrub Jay (Blue Jay) absolutely reading the riot-act to something/someone; he was in full-blown Blue Jay “scold” mode.

I stopped to look around, hoping I could see where he was. I saw Handsome and Moonshadow, standing sided by side. McKee was back in the distance. I looked in the trees and shrubs for the source, then looked back at Handsome.

It took me a second, but I finally realized that the noise maker was sitting on the hind end of my little horse. The jay was turning his head back and forth, shrieking instructions or something to both horses. I froze to watch.

Normally, Handsome is a bit reactive to new things, so I was amused when he seemed to be taking this all in stride. I’ve also seen birds sit on the horses before, so this scene was not totally new to me. Then the blue jay surprised me; he took aim with his beak and just poked Handsome on the butt. I couldn’t believe it, Handsome didn’t even budge.

I began to to grab my cell phone in hopes of catching this moment for sharing. My movement caught the eye of the jay, who promptly squawked and took flight. I wish I could have captured a quick video clip, but hopefully the description put a smile on your face.

It still amazes me how two such different species seem to work things out in such an interesting way.

Posted in Uncategorized | Leave a comment

How we view the “fallout” of training

I am very glad to see that people are increasing their observational skills so that they can see and notice when a horse is frustrated by the training process. This is amazingly good news for the horse.

I would like to suggest that we have watched horses being trained with negative reinforcement and we have become so accustomed to seeing the corresponding behaviors of fallout with the method; some even see it as a normal part of training. So the manifestations of a fearful or stressed horse are just accepted, but we truly don’t understand the level of stress we’re creating by the method.

For example if you have a trailer in an arena or pasture with a horse, you will see the horse automatically regulate the distance where it finds comfort. But when you add a lead rope and the human element and the techniques used to teach the horse to go into the trailer (which involves negative reinforcement and pressure) we are not surprised at all to see the horse backing away or showing the whites of his eyes or actually trying to escape.

Yet in my eyes these are all signs of a horse who is over threshold at the hands of a trainer using a technique.

The pressure tools in the hands of the different trainer might yield a different result. You might see a horse being asked for a single step at which point you see the release.  In this example another group of people might say that it’s better training. However, if you took the halter off and gave the horse the choice to participate, will the horse still be anywhere in the vicinity of the trailer?

So the fallout of and manifestations of the trainer’s tools are still clearly seen in how the horse immediately responds to the item once the trainer takes away all tools of coercion.

Maybe the question we need to ask ourselves is how the fallout manifest in a misapplication of a technique and how do we avoid those pitfalls?

My thought is as we get better acquainted with the behaviors, then mark the fallout of BOTH quadrants of training, we can learn to be better trainers.

Posted in Uncategorized | Tagged | 1 Comment

How the past affects our cues

I’m currently running an online course on cues and something happened today with McKee and Handsome that was so timely I just had to share.
When I performed the dance routine “Gotta Go Feed the Horses” with the minis, I had a fun section of the routine where the boys would perform two spins. Both horses would turn away from me; McKee would turn clockwise and Handsome would turn counterclockwise. We practiced these spins hundreds of times for our routine.
(If you want to see the routine, follow this link)

A precursor to this story is that I the day before this post, I dressed up in a disguise to make a point about cues and generalization and offered some cues to McKee. One of the cues I asked for was the spin. Although my disguise was a bit confusing, he certainly did perform the behavior two or three times during that session.

So today I decided to take both Handsome and McKee to an area of the pasture in which they haven’t played for probably a year. However, years ago, this area is where we practiced our routine for months and months.

I wasn’t planning on practicing the routine, but as we entered the familiar area, after couple of minutes of romping around, both boys came back and lined up in the dance formation.

Sure enough McKee started to initiate a spin, un-cued, and Handsome, being the good follower that he is, followed suit. In the routine we only did two spins, but today, they just kept spinning and spinning and in wonderful unison as I walked along.

I did not click and treat them, and it killed me ignore this lovely effort after such a long an absence from the behavior, so I did ask for a “whoa” to which they responded. That was a clickable event.

Next I gave the cue for the spin two more times and clicked and treated both efforts.

What fascinates me is, after all of this time, the amazing strength of the contextual cues. By contextual cues I mean the cues that were a part of the surrounding environment and became obvious triggers for the behaviors.  Despite the fact we have not regularly rehearsed this behavior, nor have they been in the area where we practiced, those old environmental cues were still very salient.

It also made me think again about how this dance routine was trained. The ingredients were: No pressure, all shaped behaviors, lots of careful practice and a generous supply of food reinforcers to make the game worth with their effort.

There is NO doubt in my mind that the behaviors were perceived as fun, interesting and worth playing. After all, look at the response after all of this time.

It really does provide a peek into how our horses learn in an setting that includes lots of positive reinforcers, and just how important the environment plays into what they are learning.

Posted in Uncategorized | Leave a comment

Creative thinking in horses

Probably most of you who know me, know I have a clever little horse named McKee. His antics always make me laugh and today was no different.

I was going to teach a lesson to a young boy. I was planning on using the miniature horses to show him some aspects of horse body language and how his human movements were a really important factor in the horses’ ability to follow him.

I had some pellets in a pouch and headed up to the area where McKee, Handsome and Moonshadow stay during the day, which is also a lesson and training area. The young boy brought some carrots, a VERY high value reinforcer for the horses. Closely accompanied by McKee and the student, I walked over to a small storage shed, opened the door, put the pouch with pellets and carrots on the shelf and shut the door. I also pulled a blue barrel in front of the door to block it, after all, McKee had just watched where I had hidden the goodies.

I gave McKee a handful of food and went to begin our lesson with Moonshadow. The lad and I were interacting with Moonie, exploring how horses react to scratching, petting and brushing. I love introducing kids to horses in this manner.

I heard a funny noise, and looked to see the barrel, now on its side, slowly rolling away from the shed. The door was not visible from my location, and I knew the door had a handle latch on it; I knew McKee hadn’t gone in the shed before, but I also had not hidden pellets or CARROTS in there before.

I pointed to the barrel and the boy and I giggled. Then we heard another sound, one that would be consistent with metal being tormented, followed by silence. Sure enough, when I got back to the shed, McKee was inside, the treat pouch and its contents were spilled on the floor, and he was triumphantly “Hoovering” as fast as he could.

No, he did not get scolded, nor did I remove him abruptly. In fact, I added to his pleasure by sweeping the pellets into a more convenient pile. After a minute I gave him a carrot, then offered the verbal cue of “back,” to which he easily and quickly responded.

Scaring him away or competing with him over the resource would just undermine the behaviors we had in place. So I invited him to comes back in for a few more bites, offered the verbal “back” (no body pressure cues, just a plain verbal suggestion) and he stepped backwards out of the shed whereupon I closed it.

You see, I taught this horse, through lots of creative thinking games, to be, well, creative and thinking. Clicker training is good for that, you know. When we allow our animals the time and un-pressured space to forge a solution to a puzzle, we both win. The horse gets smarter, the people reap the rewards by having a horse who learns to work willingly for the human.

To me this is the gold bullion found in the framework of Clicker training. My horses just get more and more clever. This sure can make my job easier for teaching all behaviors.

So how do I handle the horse who decides to become a Houdini?  I will try some good locks and remember to keep my sense of humor.

McKee figures out how to open this door.

McKee figures out how to open this door.

Posted in Uncategorized | Leave a comment

Cues and “Mythunderstandings”

Over the years, I’ve watched many people train animals.  I’ve seen all sorts of species being trained – dogs, cats, horses, birds, dolphins, and many others.  Some are professionals, some call themselves professionals, and most admit that they’re just amateurs trying to train their animal to do something that the human wants.  It might be a sit or fetch for a dog or longeing for a horse or “step up” for bird to step up onto the person’s hand.

One of the biggest mythunderstandings, though, is when to train a cue, whatever that cue might be.  The cue could be verbal (“sit”), or a hand signal, or maybe a particular whistle cue (three sharp blasts).  What it seems that many people do is try to train the cue at the same time they’re trying to train the behavior.  And that simply doesn’t work.  The behavior has to be trained, and trained to be solidly reliable, before it can be put on cue.

 

For example, I’ve seen a person with a horse walking around and moving their body towards the horse.  The person lifts their arm and their hand and gives a gesture.  They say the words, “Get it, get it, pick it up”.  They do this all at the same time.

 

The horse may have had some previous experience with a bucket and the horse may have even lifted it a time or two when it was alone and entertaining itself.  The human may have seen that and decided it was a great trick!  So, armed with a clicker and some treats, they went about this haphazard and backwards way of trying to teach the horse to fetch the bucket.

 

I absolutely love clicker training and I feel I can train a horse to do just about anything it is capable of doing with this technology.  However, I really feel sorry for the horse who knows the human is a food source, but he cannot even begin to sort out which of the twenty signals given will help him get what he wants (the reinforcement).

 

I would really like to strongly encourage you to study the art of training animals. You don’t need to become a professional; but for your horse’s sake, you do need to learn about the techniques and tools that comprise good training.

Learn about the Factors of Reinforcement:

 

  1. Timing:  When
  2. Criteria:  What
  3. Rate:  How much How often

 

If you don’t understand what to look for, then you won’t know what to click for.  If you don’t know what to click for, you’re just feeding the animal.  If you set the bar too high, too soon, you’ll just frustrate yourself and your horse.

 

Once the behavior is trained to about an 80% reliability rate, THEN and ONLY THEN can you begin to put a cue on the behavior.

You don’t expect, much less teach, a baby to walk overnight or to talk or to read or do math.  It takes time and practice (for both of you).

 

Do you horse and yourself a favor and learn the details of the technology.  Yes, “The Devil’s in the Details.”

Posted in Uncategorized | Leave a comment