The Goat Palace: Training is Accumulating Fast!
The goats are doing great. My journal notes are filled with superlatives. The basics are becoming much stronger and more reliable. Each session opens the door to a new possibility, something I can now ask for that would have been hard to get just a few days before. They are so much fun! I love quick, eager learners!
But before I get swept away with their current training, it is worth going back to the July Goat Diaries to see what the first steps of the learning were, not just for them, but for me, as well.
Just before Thanksgiving I had finished posting about Day Three of their training. (https://theclickercenterblog.com/2017/11/20/) Three days doesn’t sound like much, but that was fourteen training sessions. I had introduced Elyan and Pellias (E and P as I referred to them in July) to the bare bones of targeting. They had been on platforms. On day four I was planning to add in a second platform. So let me jump back to July. Hopefully, this won’t confuse you too much. I wish I had begun posting these Goat Diaries sooner so there wasn’t this overlap, but that wasn’t how my summer unfolded. And perhaps it is better this way. You can see both how these first steps can be built, and at the same time how much fun you can have using these foundation skills.
Remember in July I had not yet expanded the roof of the lean-to to create the goat palace. Instead E and P were living in the barn in one of the horse stalls. I was using the stall, plus the outside run as my training areas. So back to July . . . .
The July Goat Diaries: How We Get Behavior
My main training goals with Elyan and Pellias were: to stabilize their behavior around food so they were safe to feed treats to; and to work on basic leading. I was going to have these goats for less than two weeks. At the end of that time they would be going back into the 4-H program that Sister Mary Elizabeth ran for the children in her area. At the end of the summer the goats would be going to the county fair and to several fiber festivals. To be shown in the ring, they had to lead. So anything I could do to help them with their leading manners was a plus.
It may not seem that teaching them to target and to stand on platforms had anything to do with leading, but they are all connected. I was creating the building blocks that would make adding in a lead much easier. To help you connect the dots between these foundation skills and leading remember in clicker training there are many strategies you can use to get behavior to happen. You can lure behavior with food. I was certainly using that with the goats when I had them follow a bucket of hay back into their stall.
Food luring can be a very effective and humane management technique, especially under conditions when other skills have not yet been taught. It is much less stressful for a herd animal like these goats to follow a bucket of hay into an enclosure, than it is to be driven from behind. Getting the goats into the habit of following me and my bucket of treats was a first step towards having them stay with me on a lead.
Targeting is another way to get behavior. The goats were in the early stages of understanding targeting. I had used it to begin their introduction to clicker training.
Elyan learning about targeting.
That was step number one. The more you explore targeting, the more you discover what an incredible teaching tool it is. Targeting is very much part of leading. We usually think of targets as a visual aid. Certainly the handler becomes a visual target. But I also want the feel of the snap under a horse’s halter to become a target. In this case it becomes a tactile target – follow this feel.
Tactile targets take us to rope handling which takes us to a discussion of pressure and release of pressure. Often the mere mention of pressure makes some people cringe. That’s what they want to get away from when they clicker train. But we do put halters and leads on our animals. So the question is not do we use pressure, but how has the response to pressure been taught? Is it information or a threat?
Escalating pressure has a do-it-or-else threat embedded in it. This is what we want to get away from in clicker training. But pressure doesn’t have to become painful or frightening to have an effect. It can simply close one “door” while leaving other doors open. When you’re trying to figure out how to use pressure in a learner-friendly way, that can be a helpful metaphor.
Used well a lead provides clues that help an animal get to his reinforcement faster. Suppose I want my learner to back up. I could simply wait until I see a shift of balance back. If I’m lucky, the animal will shift back quickly, but he’s just as likely to try other directions first. That introduces more “noise” into the process.
Think about situations in your own life where having some boundaries was helpful. Computers offer us so many good examples. You want something to change on your screen, but nothing is happening, so you start hitting buttons. Is it this combination or this one? When you finally do get the response that you wanted, do you remember what you did? Can you repeat it without first trying all the errors? Probably not. How do you feel? Frustrated.
But now think about those times when the computer gave you a “not this way signal”. When you tried something that wasn’t going to work, you heard an error message. It sometimes takes me a couple of repetitions to realize that that ping I’m hearing is the computer telling me what I’m doing isn’t going to work, try something else. Oh, right. That door is closed.
At least the computer is communicating something. I must be hitting the wrong keys. Yes, I was pressing down the cap lock key instead of the shift key. That’s why I was getting that error message.
The error message doesn’t change. Siri doesn’t come on and start yelling at me. The computer doesn’t tell me if I don’t change my behavior and do what it wants, it will start destroying files. The computer remains non-reactive to my emotional displays of frustration. When I finally notice that I’ve been hitting the wrong key, it responds immediately by producing the result I want.
When I was trying to push through the wrong “door”, it gave me a clear message – try something else – but nothing else escalated. Good rope handling is very similar. When my animal partner learns to pay attention to the information the lead is providing, it doesn’t just close doors, it shows him which ones are open. What is the fastest path to the click and treat? Leads provide boundaries. Used well, they also provide very welcome information.
The lead provides simple messages. Slide down the lead and you are saying: “I want something.” Staying on the lead closes doors. Now you’re saying: “Not this way, but keep trying. There is an open door, and I know you can find it.” Releasing the lead says: “Great! You just found the answer!”
All of this has to be taught. I can’t expect my learner to understand the cues a lead can provide first time out of the box. If he’s had confusing, inconsistent, or punitive experiences with the lead, then the teaching process becomes even more involved. I’m not working with a clean slate. I have to show him through my actions that I’m not intentionally going to use the lead to hurt or scare him.
An animal that has not been carefully introduced to leads may not understand this. His learning history may tell him to try to push on the “door”. Bang on it hard enough and it will open! Goats certainly know about pushing through things! And so do many horses.
I want to build my training steps systematically so my learner can safely, comfortably discover that pushing on the door isn’t needed. When he encounters a closed door, that’s a hint. It means try a different direction. The faster you stop banging on that door, the faster you’ll find the one that is open – click and treat.
Elyan and Pellias both wore collars, but so far I had avoided putting leads on them. I wanted to give them some other skills first which would help them understand how leads worked. We were heading to leading, but not directly. The training principle is: Never start with your goal. The more steps you put between where you are and where you want to be, the smoother and more successful the learning experience will be.
More steps in part means learning to use more than one teaching strategy. So here is another training principle: There is ALWAYS more than one way to teach any behavior. The more ways I come up with to teach the same thing, the stronger that base behavior will be.
So another teaching strategy I use is referred to as free shaping. Here you are not using any prompts such as a target to trigger the behavior. Instead you are simply observing the individual and marking those moments that take you in the direction of your shaping goals.
When people talk about the magic of clicker training, they are referring to freeshaping. Yes, it is good science, but it does look quite magical when an animal begins to consistently offer a complex behavior and the handler has “done nothing” but click and reinforce tiny stair steps towards the desired behavior. There have been no targets and certainly no whips. You haven’t applied pressure by moving into the animal’s space. You’ve just sat in your chair, and now suddenly your animal – goat, horse, dog – is backing up twenty feet. Very neat.
I have always considered free shaping to be an advanced skill for both the handler and the animal learner. A handler who is just learning how to change behavior through incremental steps will miss clickable moments. The criteria will be unclear. The timing will be off. The result: a learner who is becoming increasingly frustrated and confused. A confused learner leads to a confused handler. Put those two things together and you get a mess. That’s no way to begin with clicker training.
Freeshaping may be an advanced skills, but you need to practice free shaping in order to build your skills. Here’s the mantra: for every complex behavior you teach, there will be some element that is free shaped.
I may use my rope handling skills to get a horse to step onto a mat. Once he’s standing there, I’ll free shape his head orientation.
I was going to use this concept with the goats. I got them to the platforms with the target. Once they were on the platform I wanted to free shape head orientation. My starting point was a goat who was indeed standing with all four feet solidly on the mat, but his head was reaching up towards my pockets. I knew what I didn’t want. I didn’t want him straining up towards me, or the opposite – curling his neck down so he looked as though he was about to ram something. I wanted him standing all four feet on the platform, with his head up, and looking straight ahead.
The problem was the goats never really presented me with what I wanted. They looked off to the side, or up at my pockets, but rarely were they looking straight ahead. If I insisted on perfection, my rates of reinforcement would drop. I’d get a frustrated goat, and I’d already seen what frustrated goats do. Jumping up on me was not an answer I wanted them to be practicing.
To help prime the pump I had been using the food delivery to approximate the behavior I wanted. My concern was I might be getting too much of a curl of the neck. I didn’t want to trigger head butting. So that was my question as I began the morning session. What had these goats learned from the previous day’s training? Good things I wanted? Or would I be left with “Oh dear, let me go have another cup of tea and rethink where we are.” I was about to find out.
Session 1: 8 am with P.
I wanted to make the target more meaningful to P. He clearly liked being up on his platform. So perhaps if I set out two platforms and used the target to move him from one to the other, he would begin to have a better understanding of targets. Targets are things you orient to get to other good things.
I set out two platforms, the original foam platform and a new one made out of two heavy blocks of wood. P went directly to the foam platform, click and treat. I worked on his head orientation. Mostly he was stretching his nose out towards me. I tried to catch moments when his head was down, but I needed to be careful with that. I didn’t want to teach him to lower his head into head butting position.
I used the target to move him to the second platform. He definitely got the idea of moving from one platform to another, and he was staying on the platform well. It seemed as though this was going to be a useful approach for him.
I did not film this session because there was a light rain so I have no pictures to share.
E’s session – I worked E in his stall. That seemed easier than switching the goats. I already had the makings of two platforms. I dismantled their corner platform and used two of the blue blocks as bases for single platforms. E was concerned with them at first so I put the plywood on them, and he was fine about getting up on the blocks. I again added in the scratching after feeding so he got very soft-eyed and dreamy. I liked this association. Clicks are followed by treats (exciting!) which are followed by head scratching (dreamy).
The whole peanuts took too long to eat, so I had been breaking them up. He wasn’t particularly interested in the hulls, but he did like the peanuts. I had also added sunflower seeds to the mix in my pocket, and those he really liked! We had a lovely session going from platform to platform. He was getting treats and attention. And I was getting more good data to record in my journal. Win-win for both of us.
When we were finished, I opened their stall door so he could go out into the pen with P. Instead of staying out, P came into the stall and got up on a platform. So E came back in as well.
They started sparing over who got the platform. I managed to get each one on his own platform and reinforced them for staying put. Once I have taught them individually about platforms, this will definitely be a usable approach for teaching them to work as a pair.
When I was all done, I spent a few minutes scratching them both, then I left them with some treats scattered over the floor.
The Goat Palace: Working in Pairs
So now I’m going to jump forward to the present. I just described the very bare bones beginning of using multiple platforms to work the goats together. I’ve been building on these skills both with Elyan and Pellias, and Thanzi and Trixie. It is key to being able to reduce the competition over food.
I was so impressed with Trixie and Thanzi yesterday. I’ve been working them in their pen. Each goat can now stay at her own station (a stack of plywood mats). I can move to Trixie, offer her a target to touch, click and drop treats in her bucket while Thanzi stays on her platform. Then I can go to Thanzi, and Trixie stays put. That is such a change from the dashing from bucket to bucket that we started with.
Yesterday I took them into the hallway. The narrow platforms were set out side by side. They got themselves sorted, one on each platform. I was pleased with the progress Trixie, in particular is making. Thanzi, I know will leap eagerly onto a platform. Trixie has been slower at figuring out that going to platforms is a great way to get clicked. But there they both were each on her own platform.
I stood in front of them and waited for both of them to take their noses away from my pockets. They could do it! Click, treat. And when I fed them, they stayed each in her own space to get the treat. They didn’t try to crowd in and snatch treats from one another. That’s huge progress, but wait it gets better!
Remember these goats were side by side. The treat bowls were right in front of their platforms. I could click one, drop treats in her bucket, and the other goat would stay put! Of course, she got clicked and reinforced for staying on her platform. Win-win for everyone.
Pellias and Elyan are becoming increasingly solid working as a pair. I can now consistently use their stationing behavior as a management tool. When I want them to go back into their pen, I call them and they both come running. They dash onto their platforms: Elyan on the balance beam of a thick piece of wood, and Pellias on a stack of plywood mats. Click – treat both several times. Then click, drop treats and leave.
They stay at their stations hunting for the dropped treats in the hay instead of swooping in trying to get what the other one has. That gives me time to call Galahad in and give him treats at the other end of the pen. This core foundation skill is creating much more peaceful living conditions for everyone.
Before I can move on to teaching the “fancy” stuff, first there are these basics – the universals of day to day handling. Done well, the basics become “fancy”. They are certainly fun to teach. Every day I feel like a small child who has been given another bag of leggo blocks to play with. I can build so much more with the behaviors the goats are learning! What’s next? The goats will always tell me.
Coming Next: Goat Diaries – Day 4 Learning About Goats