My horse Lukas is a seventeen year old Thoroughbred ex-racer and former rescue. He ran in three races as a two-year-old, injured his legs and left the track. He subsequently changed homes several times and was found starving in a yard and saved by a neighbor. After his rescue, I purchased him as an "Inexperienced project horse." He had many problems including bucking and spooking and required much patience, but before long his intense focus and driving curiosity began to blossom. After teaching him a broad foundation of liberty (free/loose) movements and tricks, I began to experiment with cognitive and perceptual tasks. When I unfolded his desk, he whinnied, and seeing his toys brought out nickering galore. How much could a horse learn? I decided to find out.

Current Research/Alternatives

According to most animal intelligence ranking scales, equine statistics are dismal: horses rank anywhere from fifth to ninth in intelligence comparisons between species. In addition, the equine population is thought to be a generally reactive group at the mercy of flight instincts and walnut-sized brains. I suspected that the commonly used repetitive machine trials to assess learning capabilities were missing some important components: a social, interactive element, voice prompts (particularly intermediary/leading markers) and reinforcement variations. Furthermore, I decided that the prevailing methods of force training were inadequate and even counter-productive. I decided that Lukas' lessons would resemble those that we use for children: fun, gentle and a mutual exploration into possibilities.

Techniques

My approach consists of combining three elements to create a powerful and effective training system: shaping, clicker training (substituting whistles and my voice for the click) and positive reinforcement. Shaping is the overall process, the basic foundation. The shaping strategies that I use include capturing (marking and rewarding behaviors that the student offers naturally) and creating behaviors (using cues to elicit desired responses incrementally). The click, having been previously paired with a primary reinforcement - food - to "charge" it, is an indicator that a desirable behavior has occurred. This marks the exact moment of the learner's maximum effort/performance given current abilities/understanding. Positive reinforcement energizes the methods and is given on variable schedules to maintain eagerness. Initially, treats are given steadily, then switched to a intermittent routine (i.e. every second or fourth acceptable response) and gradually shift to a random pay-off. This unpredictability of the treat regimen keeps the animal guessing and trying.

In my experience, there are several disadvantages to conventional clicker training: the inconvenience of having to carry something around in my hands, especially when riding, and the time lapse between the behavior and the click which makes it less effective. Also, during clicker sessions there are many lost opportunities for training as the animal frantically and randomly searches in a solitary manner for a behavior to be clicked for. To lead the subject to my target behavior and involve myself in the process, I employ intermediary markers to give the student hints about direction and to link attempts to the desired outcome. This is similar to the hot-cold game we played as kids: a drawn out "aaahh" or "yessss" to indicate a wanted effort or a "uh-uh" to relay a message about an undesirable response.

Training

To maximize the benefits of the lessons, I use no equipment whatsoever (including a whip). All of our activities are freely and jointly engaged in: choice and movement have given him the ability to figure things out on his own, be much attuned to me, and have also caused me to be more aware of my movements and body position. This freedom promotes creativity and initiative, yet he is extremely receptive to re-direction and the prospect of a new game. Creating and capturing are very powerful tools and can double the training results. I can capture (with whistles or my voice) any offered behaviors and then build on them (creating) to fashion elaborate behavioral chains. Lukas' tricks in themselves are not especially significant - his eagerness, comprehension and anticipation are what make his abilities meaningful and memorable.

Sessions are brief and pleasant, horses need time to absorb new information and the latitude to explore options, and I encourage this. Associating lessons with enjoyment produces quicker learning, better retention and greater generalizations. In this way, lessons tend to pick up where previous ones ended and is reinforcing not only for the student but for the teacher as well.

Results

Lukas' repertoire to date includes the smile, pose, yes, no, kiss, fetch, being blindfolded, catch, yawn, wave, pedestal work, Spanish Walk (high step), stay and come, sit, jambette (three-legged pivot), curtsey, passage (similar to skipping), bow, crossed front legs, lay down, feet all together, hide-and-seek and rear. Much of his acclaim is due to his cognitive abilities: identifying letters, numbers and shapes, discriminating colors, and his grasp of proportion, same/different, spatial relationships, object permanence and absentness. Lukas has been on NBC, CBS, ABC, CNN, HLN and Inside Edition and in feature stories released by the Associated Press and America Online. Newspapers, forums, blogs and newsletters world-wide have shared his story and his journey has been heard on Horse Talk Radio, Pet Talk Live Radio, Pet Place Radio, All Paws Pet Talk Radio and RFD-Radio among others. The World Records Academy recognized Lukas as the "World's Smartest Horse," and Guinness has approved his record "Most numbers identified by a horse in one minute: 19." Also, Lukas was nominated for the 2010 Equine Vision Award sponsored by Pfizer and American Horse Publications. He is the Spokeshorse for After the Finish Line, an organization dedicated to finding homes for ex-racehorses and he's been invited to participate in the Wounded Warriors program.