Clicker Training the Goats

I’m taking an introductory animal behavior course right now, on line; part of my quest to learn as much as I possibly can learn about this fascinating field. So I’m studying the theory of how animals (ourselves included) learn. I’m parsing the jargon of behaviors and reinforcers and punishers and extinction and antecedents and consequences. Of establishing operations and abolishing operations; of discriminating stimuli and delta stimuli; of topography and function.

At the same time, I’m jumping into the fray and practicing–learning by doing. Once I learned about operant conditioning theory last year, I couldn’t wait to start experimenting with as many different animals as possible. Target training is a gratifying place to start; it’s amazing how quickly an animal at liberty finds agency, making the environment work in their favor. The animal stumbles onto the conclusion that touching a target brings treats, and it will keep searching out that target as long as treats are motivating. But it’s one thing to read about targeting; it’s even better to get out there and try it.

Good work, goats.

My very first learner was my dog, incredibly food motivated. Soon enough I had her working off scent to target my keys, finding where I’d hidden them among the couch cushions. Then I worked with my “crazy” thoroughbred Luna, and noticed that she remained calm and quiet throughout entire training sessions. That was an eye-opener. Thus convinced of its utility with horses, I started bringing target training into my veterinary work. Then, I got goats.

Did I know much about goats before training them? Not really. From veterinary school I remembered they had 4 stomachs and got in trouble with parasites and need supplemental copper, but I knew next to nothing about goat personalities. I did have an inkling about their appetites, so I borrowed a couple of goats to eat the invasive blackberry vines out of my back pasture. Now, I am training the goats.

Like Agnes the mule, these goats need some hoof care, although unlike Agnes they do not need it urgently and are not generally averse to human contact. These four year old, female, Boer goats seek out human contact. They are from a friend’s 4-H herd, socialized to humans and eager to see what we’re up to.

Also unlike Agnes, these goats are small enough that a strong adult human can restrain them and force compliance with necessary things like hoof care. But come on–they’re friendly goats! I see pictures of goats trained to pull carts in harness; why not teach them something more simple, like how to stand willingly for hoof care?

Out comes the target. I imagine a future in which these goats stand calmly while I run a hand down each leg; in which they lift up one hoof and balance on three others; in which trimming their hooves will not require me to call over a strong guy to come hold them still.

The goats–quirky, observant characters–naturally glance over at the new object I am holding, this bizarre white blob on a stick. I mark that glance with a click! and offer some alfalfa pellets. YUM–they like that, and they see where the pellets come from. They nose around the feed pouch at my waist a bit while I hold it closed, non-reinforcing. Soon, another glance at the target–click! Food reward. It doesn’t take long for the goats to make the switch from randomly encountering the target, to deliberating checking it out. It’s the only route to those blessed alfalfa pellets, and they’re not stupid.

In fact they’re remarkably fast at this target training exercise, at least as fast as my dog was. I repeat several days in a row, and it isn’t too many days–maybe 5–until I am stationing the target at the fence and standing 10 feet away. The goats nose my pouch to no avail, then go touch the target in order to hear that click!, and get their treat.

It’s time to move on to hoof handling … (to be continued in future posts.Check back often!)

Thinking horses, or maybe GOATS? Think positive!

When Positive Reinforcement Training Goes Bad

It’s easy for any training method to be misused. In the case of positive reinforcement training, we start to lose good results when the training becomes more about the treat than the behavior.

I’m working with an Alaskan Malamute at the moment named “Klondike”. He’s a juvenile (20 month) intact male with a behavior complaint of excessive jumping, on people, which quickly escalates to ever more aggressive posturing. He has scared a handful of animal people (neighbor, farrier, vet) to the point that some have mentioned he might need to be put down.

Klondike, learning to be brave. This was our second session together. He’s not jumping on me, but he’s still stressed.

Klondike has been put on prozac by another DVM and the owners were taught to train with treats–ad nauseum–for good behaviors. After I started working with the dog, the owners said that I’m the first person from whom they’ve heard anything about phasing out the treats, or about rewarding the dog with attention/petting instead of (or in addition to) the food rewards.

The wife mentioned that she didn’t like all the emphasis that had been placed on giving treats, and I asked why. “It’s like giving a sticker to every kid in kindergarten. Everyone just gets told how good they are all the time and they don’t even have to do anything!” she quipped. I would paraphrase her objection as being a stand against grade inflation: use rewards injudiciously, and pretty soon we have an animal (child) who expects to be rewarded for little effort. The animal (child) doesn’t even really know why they’re getting rewarded, they just learn to expect a reward whenever the trainer/owner (teacher/parent) is around.

This is all very well when we are working with a fearful subject, animal or child, and need some way to establish a pleasant association with our presence.

But if any real training is to be done, rewarding with food treats needs to be done precisely.

Let’s look at the flip-side of this: If an animal (child) doesn’t understand why the rewards show up, it follows that they’ll be equally unaware of why the rewards stop. The animal’s (child’s) behavior begins to be shaped by an addiction to the possibility of a reward, instead of being shaped by understanding that certain behaviors make life more rewarding.

That is positive reinforcement gone bad. Instead of a cooperative, pleasant, or useful partner, we have a spoiled maniac obsessed with getting their next fix, and frustrated when that fix doesn’t show up. I see this every time I see an owner dangle a treat in front of their pet as they ask for a behavior; to be honest, it kind of makes me nauseous. Watching an addict in action, animal or human, is no fun.

We’re operating at a neuro-chemical level when we use food treats–it’s fundamental evolutionary biology which dictates that animals are hard-wired to want food. Once animals associate a certain behavior with increasing the likelihood of food, they will reliably perform that behavior. But it’s not easy for a trainer to deliver food at the right moment in time. We have to dig for food in our pockets, pick up a small piece, and hand it over. During all that time (we’re talking seconds) the animal has probably changed its behavior. Maybe it was doing what you wanted when you started to reach for the treat–in Klondike’s case, let’s say he was walking near your left side, at “heel” position. By the time you get the treat from your pocket, Klondike has wandered a little ahead or behind, or off to the side, or found a squirrel to chase. The training moment is lost, but you now have a treat ready, which Klondike smells, and starts begging for. Klondike will try a host of behaviors in order to get the treat, but Klondike has no association between what he did a few seconds ago that you actually liked (walk at your left side) and having the treat show up.

Unless we can precisely pair a desired behavior (heel) with a reward (food) then we’re not training. We’re just casually carrying around one of the most powerful motivators known to animal-kind. The bottom line is that food is a very, very powerful motivator. As with any power tool, food tends to take over the project unless it is used with great precision.

Great precision is really hard to master, and I think this is why training with food can get a bad wrap. The professionals working with Klondike before me were not wrong to introduce food to Klondike’s behavior modification. But they were misunderstood. They were explaining how to use food to diffuse a tense situation (Klondike around new people). That is very different from using food to reinforce specific behaviors (like teaching “heel”.)

Stepping in to say that they could use food less often, or use attention instead of food as the reward, was my way of saying, “Let’s not have food around regardless, while you’re teaching a specific behavior. Let’s teach Klondike that what he does can make the food show up.”

No Exit

One of my best teachers is a miniature, adolescent mule. Apparently her childhood was traumatic, because she came out of it with a solid distrust of all things human.

Agnes with Daisies. (Donkey “Daisy” is wearing the fly mask.)

Her name is Agnes. She is a resident of Zeb’s Wish Equine Sanctuary, where she arrived after Oregon Humane Society seizure due to neglect. At approximately 6 years of age, she and her 8-inches-too-long, ski-tipped hooves leaped from a horse trailer onto the Sanctuary, poised for transformation.

Two years and counting, we’re still at an impasse with Agnes; hoof care without incident remains an elusive goal. “Incident,” at the most recent trim, involves IM sedation administered by air-powered dart gun, embellished with halter and lead rope snubbed to a solid post. With chemical assistance, in other words, we can get Agnes slowed down enough to deny her all chance of exit.

And then she has to suck it up and deal with us, sedated though she may be. This may not be the most pleasant experience for Agnes, but hoof trims are not optional. After all, overlong feet was the hallmark of her Sanctuary arrival in the first place.

The end therefore justifying the means, Agnes and her care-giving team collectively tolerate the use of chemical and/or physical restraint, in order to preserve soundness. No hoof, no horse–uh, mule.

But ends justifying means is a slippery slope, perhaps one of the slipperiest slopes in the landscape of communication. I slipped from critical hoof care justifying no-exit handling, to my own ignorance justifying it as a training method.

It sounded something like this, when I explained it to Jody and Megan of Train With Trust: “So I worked on desensitizing Agnes to human touch while she was sedated, after her trim. Unless we deny her a flight path, we just can’t get near her.”

“You did what now?” Megan and Jody responded, via Zoom call. Zoom turning the screw of awkwardness because all parties actually got to watch me wriggle and squirm further into my hole of logical fallacy; I should have turned off the camera.

But you know what a good teacher does? They ask questions. Megan and Jody are good teachers. Therefore, Megan and Jody asked. And let me wriggle, and then helped.

“So there’s this whole issue of consent …” they began, and as I listened I began to learn. If the animal is not allowed choice, if the animal is not allowed the option of leaving, then we are not training humanely. We are capturing, forcing, teaching the animal that they have no choice. We are facilitating learned helplessness.

There are so many ways that this “consent” fallacy plays out within our own species, date-rape being an especially abhorrent one. The paradigm is that X–mistaken or not–believes they cannot leave, so Y believes that X wants to stay. Y is a terrible, terrible teacher. Y never bothers to ask, “Do you want to be here and do this thing?” and X is scarred for life.

Whew. There is nothing quite like being tactfully compared to a date-rapist to shock one into a new perspective.

So thank you, Train With Trust. Thank you, Agnes, for looking at me with those large, gorgeous eyes, and asking me, “Can’t you think of something different?” Yes, Agnes. I can do better. I’m learning; and I thank God for that.

–Thinking horses? Think Positive

Stuck in the Trailer, Totally Free

Here’s a new twist on a perennial horse problem: Finn the 4 year old Fjord gelding refuses to get out of his trailer.

Finn, at 19 hours stuck.

You’ve read in these posts how target training helped Mr. Ed, a horse who refused to load up. His trailer issue is much more common. And while I’ve seen plenty of horses exit trailers with something less than grace (my own mare included), I had yet to find a horse who flat refused to unload.

Until I met Finn.

Still tender in years, Finn has a well-established history of trailer-block, and his story is a great one to illustrate the power of positive-reinforcement training. As with trying to get Mr. Ed in, many methods of getting Finn out prove futile, some dangerously so. Target training brings Finn out safely, at liberty, within an hour or two.

An hour or two? What’s the use of that? Consider: Finn is a 1,000 pound creature who would rather go hungry than leave a trailer; even a two-hour solution is better than none.

He’d been stuck in the trailer for 19 hours by the time I saw Finn. 19 hours, without food or water, too afraid to get out. Food and water had been left as lures outside the trailer, but seeing them there was not enough to convince Finn to take the big step down.

That’s stark evidence for how “stuck” this horse was–and it’s a critical time for me to interject this: DO NOT withhold food or water from your horse. Once more, with feeling:

DO NOT WITHHOLD FOOD OR WATER FROM YOUR HORSE. The consequences could be fatal. 

–I’m a pretty laid-back vet and I don’t say “fatal” about many horse-owning mistakes.

But we all make mistakes, until we know better (heck sometimes even after we know better, who am I kidding?) Anyway, an important digression.

Mistakes aside, the fact that food alone would not get Finn out illustrates a fascinating facet of clicker-training: successful training with treats is not about the food. It’s about giving the animal space to problem solve, with food used as the reward.

I like this because clicker-training is often dismissed with its reliance on food; I was once a disparager myself. Training with food is perceived as less “real” than training with praise alone, not to mention with pressure/release alone. But let me say this: all good training, regardless of method, involves offering something of value to the trainee. We flatter ourselves when we decide that simply being with us should hold enough value. Or simply finding a release from pressure.

Want to hear how well adding pressure worked for Finn? What I’m about to describe is not about using pressure correctly–instead it’s about desperation (and a heck of a lot of good luck that no one got hurt.) I have no doubt that a trainer skilled in pressure-and-release could eventually get Finn out of a trailer. It would probably take about two hours, maybe more. Who knows? As any good trainer will tell you, it will take as much time as Finn needs.

not a good surface for trailer unloading

Several months ago, Finn stepped forward out of his step-up trailer onto asphalt, slipped, and fell. I wasn’t there to see it, I don’t know exactly what happened, but most likely it was scary. Most likely it left an impression of risk for the fella. Thereafter, it has been a struggle to get Finn to unload.

Thereafter, folks have tried pressure and not a whole lot of release. Pressure has been a lunge whip tapping on his butt. Then tapping a little harder; then “tapping” really, really hard.

Someone added a butt rope, and tried to pull Finn out. Backing out was tried, but Finn had no more understanding of this than going forwards. How has Finn gotten out? With a rope on each front leg, as well as a butt rope, and brute force to pull his front legs forward and drag him out. The last time Finn unloaded, it was with a butt rope attached to a tractor, and the tractor did the pulling.

I. Can’t. Even. How did nobody get injured through all this?

Injured or not, and no I wasn’t there to see any of this, I can assure you Finn got stressed. Any of those pressure-added methods would have left Finn with elevated stress hormones, muscles tense, nostrils distended, eyes wide. As mentioned, none of these attempts sound like a correct use of pressure, since none of them included the critical release portion which allows for learning. That’s how pressure-and-release training goes bad, by forgetting about the release (as well as by ramping up the pressure enough to risk physical harm.)

That’s also where positive-reinforcement training really shines, when you have a horse who is scared enough that it’s almost impossible for them to find release. Instead of relying on pressure, positive reinforcement training relies on liberty, and asking the horse if they can figure out how to get our reward.

When I found Finn at his trailer edge at the 19th hour, I had prior knowledge that one of his handlers had already taught him to find a target. I got my target, and sure enough Finn reached his nose to it with alacrity. Then, instead of trying to get Finn “out of the trailer” I focused on increments of forward motion. I held the target slightly out of reach, and Finn leaned forwards for it. Click for the lean. Next time, he leaned forward and moved a front foot also–I clicked for the step. Every time Finn did anything which had him trying to move forward against his imaginary barrier, he got clicked and reinforced.

No ropes, all smiles: Sammi and Finn after less than an hour of work to get off the trailer.

I was actually at the farm for routine veterinary work, so I had to move on to attend to that while Sammi Jones finished the unloading process. Sammi is the one who had taught Finn how to target, and Sammi is the one who has gotten Finn unloaded in the past, without force.

Finn is still practicing, and in time he’ll step on and off a trailer with aplomb, routinely. For now, we have a traumatic history, a stuck horse, and a method he loves which shows him he can do this.

-Thinking horses? Think positive.

Covid-19

Isn’t it nice to know that what you do really matters?

That was my first brave response to realizing that Covid-19 was here among us, and dangerous.

Image result for covid 19 image
A model of the Covid-19 virus

My ignorant response had been different. By about March 10th, my peers were tossing around the term “social distancing”–epidemiological jargon for get the heck out of my space. But I didn’t understand why. The Wuhan virus, as I knew it then, was a respiratory threat with an epicenter in China, and yeah, it was making its way around the globe thanks to international travelers. But from what I’d heard the illness wasn’t that big a deal. Most of us would develop mild symptoms, if any at all, and the death rate was lower than that of seasonal influenza. We don’t panic about the flu every year, I reasoned–why are people raising panic about Covid-19?

So on March 12th I called my friend, a physician with the VA in Montana and my go-to on medical questions, to ask why all the fuss? “It’s not even as bad as the flu,” I pointed out.

“Yes, it is,” she countered. “It’s worse.”

She told me to start reading the reports from northern Italy. She taught me concepts like “flatten the curve” and had me watch a video from CBS Boston where her cousin, head of infectious disease at Brigham and Women’s hospital, explained it clearly: social distancing was not about us, it was about preventing a melt down in our medical system.

The gravity of a Covid-19 pandemic started to sink in. I had been focused on the idea that as a naive population faced with a novel virus, our best defense against the disease would be the acquisition of herd immunity. As a large animal veterinarian I had been trained to think about herd immunity. With the mortality rate relatively low, why not let the infection spread through the healthy population and stimulate natural antibody production, while we protect the elderly and otherwise immune-compromised? We don’t have a vaccine for this disease yet, so outside of hand-washing, only natural immunity will prevent people from spreading disease.

(A few days later, on March 15th, I felt somewhat vindicated in my logic after hearing it echoed by the chief science adviser to the United Kingdom.)

Herd immunity is all well and good as a scientific principle, my friend acknowledged, but even a low mortality rate causes high numbers of dead in an urban population. Public health officials weren’t claiming that we could prevent people from getting sick, she continued to explain. Public health officials were stressing the importance of having fewer numbers sick all at one time. Flattening the curve means that the same numbers of people become ill, but at a slower rate.

There is a rate of infection at which our hospitals simply can’t keep up. Hence the warnings from public health officials; hence the advice to read and learn about what was happening in Italy. Northern Italy, a modern, western, wealthy economy, was overwhelmed with sick and dying patients. Doctors were having to ration ventilators, choosing who to treat and who to let die. Bodies were piling up in empty churches, backlogged on their way to crematoriums. If it could happen it Italy, it could happen here in the wealthy, invincible US of A.

I didn’t feel brave about this from the start. At the grocery store, faced with bare shelves which posted rationing notices in lieu of toilet paper, I actually felt scared. I remembered being a little kid and trying to understand my father’s explanations of beleaguered economies. He had been an international reporter for the Wall Street Journal, witnessing world economies first hand. I vaguely understood from him that other people, in other countries, had to line up for limited supplies of staples like bread and milk. But here I was looking at bare shelves in my home town; I’d never experienced this before in my lifetime. I’d never thought it was even possible to experience this in America, near a major city: there was no toilet paper to be had.

When I came to the store a few days later, scarcity had spread to the bread and milk aisles. And still no toilet paper.

I ended up accepting a few rolls from a friend. She handed them to me, in the sunshine outside her country home, reaching over from about six feet away. We were social distancing.

Why? Because it’s something we could do. Watching our economy crumble around us as a result of trying to slow the viral spread, it’s easy to feel helpless. But what epidemiology tells us about social distancing is that, if everyone participates, it works. Each and every individual one of us can actually make a difference in the spread of respiratory infection.

We have to embrace this concept, and we have to act on it. So much of being human is tempered with self-doubt, with wondering why we are here, or whether our measly little self can make an impact for the greater good. The Covid-19 virus is giving us a wake-up call: what you do matters!!

Please–stay home. We’re going to get through this, together.

That’s my space, Cowboy!

I introduced Cowboy a few days ago, my latest clicker project who needs to resolve issues of crowding his humans. Another concerning behavior has been kicking at horses or people during group rides. Overall I’d sum up Cowboy’s issues as one of spatial awareness; he needs a more consistent and confident sense of where his space ends, and someone else’s begins.

Cowboy yielding space. What can you read from his expression?

It’s really tempting to guess whether we are able to increase Cowboy’s confidence. As a horse lover I can’t help but want to “know” this, but wondering about such nebulous things as emotions can get in the way of actual training. We’ll probably never know for sure if Cowboy feels confident, cocky, has high or low self-esteem–but we can guess how he feels by using objective criteria. On an every-day scale, we make these “guesstimations” by reading body language. On a micro-scale, we study neurochemistry and brain imaging to make similar, perhaps even more compelling, guesses.

I’ll save the neurochemistry part for future posts.

In the meantime, we want to know if Cowboy has stopped kicking on trail rides, and if his owner’s feet are getting stepped on any less.

Our last post ended with Cowboy getting introduced to the target. Again, target training with an exuberantly curious horse like Cowboy is a breeze, because they are bold enough to start checking out the target immediately. Within 4 or 5 clicks, Cowboy notices that every time his curiosity gets the best of him and he sniffs the target, he hears a click sound and he gets fed a treat.

Then we start moving the target around. Cowboy checks the target out in its new position and whaddya know, the same click! and treat happens. We move the target to a handful of new positions, all within a few feet, and we notice two things. First of all, Cowboy is actively seeking the target in anticipation of his treat. Second of all, as soon as he hears the click he’s looking away from the target, back to us for his treat. So he knows he can make the click happen (nose the target), and he knows that click means there’s a treat soon to follow.

Another easy thing about a horse like Cowboy is that he’s so curious that he really enjoys the game, and we can keep clicking for a fairly long time, maybe 30 clicks or more. But for most horses, if following the target around is taking quite a bit of deliberation, we might need to stop a training session after only 10 clicks. Honestly, 10 is plenty. I leave Cowboy and his owner to play with this new game for the week: Find the target, touch it, hear click and get fed.

Over the course of the first week, Cowboy plays with finding the target from a greater and greater distance. In small steps, mind you; “greater and greater” means we start with 3 feet away, then 5 feet away (target hand-held by the human). Then we put the target on a wall somewhere or stick it in the top of a traffic cone, and have Cowboy find it from 6 feet away, 10 feet, etc. We point to it each time, and Cowboy learns to follow our point and find the target.

At our second weekly training, Cowboy is finding the target in various positions around his stall. He seems a little unsure, so I show his owner how to lower the criteria a bit (make it easier) in order to increase the rate of reinforcement. See how I used that behavior jargon? In plain English, I made the target easier to find, so that Cowboy could get rewarded quickly and often, rather than struggling to get the “click” each time. Lots of fast repeats make Cowboy even more reliable about finding the target he’s pointed to.

In the next session we introduce the game of liberty leading.

~Thinking horses? Think positive!

When do they get to be right?

Cowboy, the latest of my clicker-training clients, is working with me on yielding space to approaching humans. Cowboy is curious and friendly. He is more apt to smother bipeds with attention than he is to avoid getting caught. He’s cute to the horse-initiated, but at 16.2 hands and over 1200 pounds, his exuberance is intimidating to the non-horsey, not to mention potentially dangerous.

His owner would like to have her husband, decidedly non-horsey, be able to feed Cowboy without incident. It would be nice if hubby could bring food to the stall, ask Cowboy to back up, and place the food without wondering if a horse will be in his way when he turns to leave.

Time for Cowboy to learn targeting!

On my first visit with Cowboy I take a detailed history. The question I’m left with (hypothetically) is, “When does Cowboy get to be right?” I hear stories of a horse who is so tightly herd-bound to a travel mate, at a show or trail ride, for example, that he had started to kick others who might be in between him and his chosen horse partner for the day. That got him sent to a trainer, where he stayed in a stall for the first time in his life.

Apparently he barged out of the stall and nearly ran over the trainer, who “didn’t take kindly to the situation,” whatever that means. Then if shut in his stall he made such a fuss that he was given kick chains to wear. To mitigate his kicking out at other horses, he was tied in the arena for hours while others rode by. I’m not sure how this worked (was the idea to have horses coming past him so often that he finally gave up reacting?) and the owner didn’t have a first hand account; she was simply telling me what she had heard from the trainer.

Cowboy was sent to the traditional trainer and, from what I hear, received a traditional training dose of corrections. Punished for running out of the stall (more on this in a second.) Punished for being upset in his stall. Required to stand tied for hours, unable to leave a potentially unpleasant situation. When did Cowboy get to be right?

Only when he gave in to the pressures around him.

Let me pause for a minute to address the issue of safety. SAFETY ALWAYS COMES FIRST! So in regards to whatever the trainer did when Cowboy ran out of his stall, I can appreciate that she probably made herself big and scary and intimidating, whatever she felt was necessary to keep herself safe from impact. Sometimes the moment requires this. Better, of course, would be to set the horse up for success in the first place.

Better, in other words, would be to switch our focus from all the things Cowboy is doing wrong, to what does Couwboy do well? Let’s build on those things and help him fit into the odd requirements of domestication. Requirements like yielding space to humans.

I love clicker-training for horses like Cowboy because it allows them to be right, right from the start. Cowboy is naturally curious and exuberant? Perfect. Clicker training is all about checking out what’s new in the environment. Bring a target (new) into the environment and Cowboy is absolutely going to check it out. That gets him a click and a treat.

Hmmm, thinks Cowboy. I like treats. Within minutes he’s figured out that touching the target earns him a treat, and he’s eager to follow that target wherever we place it around the stall.

In future posts I’ll detail the steps we took to bring Cowboy from in-your-face to here’s-your-space.

Thinking horses?
Think positive.

Mr. Ed, pt III

“…all of this trailer training was done at liberty. Ed was bound to participate only by his choice.

–Kath Mertens, DVM

… continued from a previous post

We left the previous Mr. Ed installment with him stepping both front feet into the trailer to reach his feed tub. But come hell or high water, Ed made it clear that he was not willing to get all four feet into the trailer; he’d rather go hungry.

Letting him “go hungry” was not an option, so I had to find another method to teach Ed that the trailer was worth stepping into. It was time to go back to target training with the clicker.

…I had him target inside the trailer without me, leaving the target perched on my hoof stand.

I started basic target training outside the trailer, first holding the target in hand, then placing it around his pasture, and rewarding him each time he reached to touch the target with his nose. Then I brought these new skills to the trailer, having him touch the target while I held it from inside the trailer. Next I had him target inside the trailer without me, leaving the target perched at the top of my portable hoof stand.

Ed was sweet and willing and enjoyed the game, but we didn’t actually make progress on getting further inside the trailer. Reach in to target? Yes. Step front feet in to target? Yes. Step a hind foot up? Nope. We were still stuck, half in and half out. Ed would step both front feet in, barely start to lift a hind hoof, and stop.

I wracked my brain for a way past this dilemma, thinking about splitting my desired behavior (trailer loading) into sequential tiny steps of success. Don’t train the end picture, I reminded myself. Train the next step. And then it was clear: if we were stuck at the hind feet, I had to train the hind feet to take a step.

Back to the pasture we went. I asked Ed to walk circles around me, and I clicked each time he moved a hind foot. This meant lots and lots of repeats of rewarding one step at a time, like this: Walk on, Ed. Ed moves front foot, hind foot, CLICK! Except I clicked at the beginning of the hind step, so the timing was more like: Ed moves front foot, hind fo–CLICK!! I tried to get my click as close to the beginning of the movement as possible.

I still needed more. I needed something to emphasize the stepping up part of trailer loading. (Weeks earlier, as this whole loading journey began, Team Ed had tried bringing over a ramp-up trailer to rule out the step as the limiting factor. To no avail: Ed was equally unwilling to climb a ramp as a step-up. He marched both front feet in, and stopped.) Conveniently, a dry-lot corner of my pasture is delineated with felled telephone poles. Originally intended to contain a layer of hog fuel within the dry-lot, the poles have also provided physical therapy for horses recovering from leg injury. Horses put in extra effort to lift each leg over the pole, passively providing range-of-motion exercise.

It dawned on me that if I brought Ed to the poles, I could click as he stepped each hind foot up and over. I’d be reinforcing not just a step forward, but specifically a step up and forward.

It worked great. I want to pause to emphasize here that all of this trailer training was done at liberty. Ed was bound to participate only by his choice. I had made a promise to always pay him for his efforts, but he was free to shrug off the work as not worth a mere treat at any time.

To the telephone poles we went, and I held the target out as an invitation for Ed to cross from one side to the other. Instead of waiting until he reached the target to grant his click, I reinforced the moment each hind foot lifted up and over the pole. Then I modified to clicking when each hind foot was at its apex over the pole.

hind leg in!

I brought the exercise back to the trailer, where we had been stymied at Ed giving up instead of completing a hind leg lift. Now he’d been primed at the telephone poles. I went back to clicking just as he began to lift a hind leg. We did it over and over and over again. His effort increased infinitesimally each time, until–plop!–he lifted a hind foot up and rested it inside the trailer.

Beautiful! I ended the session and gave Ed his jackpot payout.

4 hooves in
Well done, Ed!

As this process was playing out in real time it often felt like progress was achingly slow. Ed’s focus was interrupted several times by angry squirrels chattering in the trees overhead. Or by my dog barking at the wrong time (easy enough to put her in the house, of course.) I would wait minutes at a time for the next step into the trailer, but as long as Ed was willing to stick with me and try to figure things out, I was willing to get to a good end-point for the day.

On October 12th our end-point was targeting. On the 15th I started clicking his footsteps. On the 16th I clicked his hind feet over the telephone pole. And on the 17th, praise be, he clicked all the way up into the trailer. So basically, it took 6 days, inclusive, to take a horse from stuck at the edge of the trailer to volunteering to stand inside. No stress, no tack; no other way to train this case that I and my team could think of. Well done, Mr. Ed!

–With special thanks to: Suzi, Cindy, John, Holly, Paige … who all believed he’d get in.

The Misunderstanding of Positive Reinforcement.

I am one of a growing army of people trying to bring positive reinforcement training to the horse world. The most familiar picture of positive reinforcement training is that of clicker training. Many people have heard of clicker training with dogs; what’s less well known is how useful clicker training is with horses.

One reason I’m so passionate about bringing clicker training to horses is that I am a convert myself. I first dismissed positive reinforcement training as a treat-inspired gimmick promoted by trainers without talent. Because these trainers were not adept at motivating the animal themselves, I figured, they relied upon motivating with clicks and treats. Clicks were a crutch for trainers with poor timing. Treats were basically used as bait to get the animal to do what clumsy trainers couldn’t convince them to do otherwise.

Furthermore, I objected to the mistaken implication that “positive” reinforcement meant that it was good, better, morally superior.

This pervasive misunderstanding of the “positive” in positive reinforcement is slyly exploited in any marketing of this training method. This exploitation drives me nuts, because I think it prevents people from truly understanding what the method is all about.

Positive reinforcement is not better, but it is different. Is a hammer better at building a house than a screwdriver? No, it’s not better. It’s just different. The builder with both tools in his box will build a stronger house, faster, than the builder who needs to rely upon just one.

Positive, and negative, reinforcement are for the trainer what hammer and screwdriver are to the builder. They are two fundamental tools, and being skilled with both allows the trainer to do her job more efficiently. Efficient, effective methods mean a lot less frustration for trainer and animal alike.

So here’s the deal with the naming of reinforcement types: “Positive” and “negative” are taken from scientific jargon, with the same value proposition that is associated with naming positive or negative integers on a number line. “Positive” means something is added. “Negative” means something is subtracted. One is not better than the other, nor is it necessarily preferred by the animal.

Many “Natural Horsemanship” trainers would be appalled to learn that what they are practicing everyday, as they use pressure-and-release, the language of Equus, or whatever they want to call it–what they are using everyday is what the animal behavior scientist calls “negative reinforcement.”

What? I thought negative meant that the animal was getting punished, turned away from something?

No. Negative reinforcement means that when we see a behavior we want to see more of, we reinforce (encourage) that behavior by removing something from the equation. In the case of “Natural Horsemanship,” the thing removed is pressure.

Positive reinforcement means that when we see a behavior we want to see more of, we reinforce that behavior by adding something. Very often, this addition is a food treat.

Clicker training is just going one step further to link a secondary reinforcer to the equation. Food is the primary reinforcer. The click means nothing to the animal in and of itself, but over time we “load” the clicker with meaning by conditioning the animal to expect food after every click.

Remember Pavlov’s dogs? How Pavlov rang a bell just before feeding time every meal, and eventually the dogs would begin to salivate just from the sound of the bell? In slang behavior-speak we would say the bell became “loaded” with meaning; it began to mean “food is coming.” In proper behavior-speak, we say (as Pavlov demonstrated) that the animal was classically conditioned to expect the treat after the sound of the bell.

Once you have an animal conditioned to expect a food reward after every click, it becomes much easier to use the click to mark the desired behavior, rather than to use the food. Food is clumsy, We fumble for it in our pockets. We drop it. We take too long between seeing the behavior we like, and reinforcing that behavior with food.

Conversely, the clicker is precise. It is exact, succinct, immediate. Whether we use a hand-held clicker or make an audible click with our mouths, it is easy and repeatably fast to mark the behavior we like with the click.

Then, of course, once the click has been suitably loaded, that classical conditioning buys us a little time between saying “Yes!” to (clicking) the behavior we want, and getting that treat to the animal. We have time to be imprecise and fumble with the food, because the click has already delivered the message of exactly which behavior we liked.

It’s kind of like the difference between getting handed a paycheck for doing our work (the paycheck is the click) and then cashing that check at the bank (or depositing it into our 401k if we’re prudent!). We get the actual cash reward once we’re at the bank, but we know that we are not getting rewarded (paid) for going to the bank. We’re getting rewarded (paid) for having done the work. That’s a bit of an analogy stretch, but worth considering.

There is oodles more to be written in this blog about positive reinforcement and secondary reinforcers and how clicker training works. But for the sake of getting this post out and leaving you with a thought to chew on: positive reinforcement isn’t better–but it is fascinating!!

Thinking horses?
Think positive.

Mr. Ed, pt II

…continued from previous post

So in my effort to get Mr. Ed loaded into the horse trailer, I had begun feeding all of his meals there, trying to de-sensitize or counter-condition him to the trailer. He wouldn’t like it at first, the theory goes, but he’d be hungry and over time–first tentatively grabbing bites of food and then settling down to eat as he realized nothing bad was happening–over time he would relax and the trailer would be just another place.

… his hind feet remained outside.

This theory worked to a point. The reason it worked at all is that Ed, toothless, depends entirely upon prepared meals for his nutrition. Grazing for Ed is a soothing pastime only, not a source of calories. He wads up mouthfuls of grass between his toothless gums until finally spitting them out. When it is time to actually eat, he finds his Equine Senior Grain.

I parked the trailer at the corner gate to my pasture, leaving Ed free to graze at will, and I pushed his feed tub a little further into the trailer each day. Ed was happy to eat from the edge of the trailer, all four feet on the ground. He was willing to step both front feet in the trailer to reach his tub. With each passing day, he stretched further inside the trailer, but his hind feet remained steadfastly outside. Ed got really good at grabbing the feed tub and pulling it towards him, back to the trailer’s edge where it was easily within reach.

feed bin weighted down, tied to corner of trailer. To no avail.

“Darn it, Ed! You’re cheating,” I grumbled. I brought over a small log to block the tub in position. I brought in a 40 pound cement block to weight the tub down from within. Ed laughed at all of these, toothless gums still grabbing the bin and pulling it into a convenient position.

So I brought in a 40 pound weight, AND I tied the feed tub into the far corner of the horse trailer, forcing Ed to walk on in or go hungry.

Ed chose the hunger strike.

Let me interject here that I was acutely aware that Ed maintain healthy weight during this process. Nor would I let him go long with an empty stomach, as this increases the risk of equine stomach ulcers. So I’d monitor his behavior until it was clear he was willing to go hungry rather than change his habits. Then I’d admit defeat, and move the food back to where he could reach it.

Given that Ed was willing to go hungry until I caved, it was time for a different strategy.

“Well, have you tried target training?” asked Jody Ambrose of Train with Trust. Of course I had not, I admitted with an abashed grin over our Video Chat. That would be like the cobbler’s kid having new shoes. Or the target-happy-veterinary-horsemanship-trainer training her own horse.

“Well I’m actually not sure he can see very well …” I said, trying on an excuse. “For that matter I know he doesn’t hear very well,” which is true. Many a time I’d called out to him as I came to visit in his pasture, only to elicit a major spook once he finally saw me. Would he be able to hear a click, or see a target?

But it was so obvious this was the right thing to do. I decided to stop being hampered by an unconscious fear of failure and get out the clicker and target.

continued in Mr. Ed, pt III