Monday, January 13, 2014

How effective is the "Catch him being good" theory?

All titles available  in paperback and ebook formats
from iUniverse,
Barnes & Noble,
and autographed copies are available from the author

Waiting to capture a behavior? | Stale Cheerios Blog

My horses are probably as intelligent as any out here.  At least most of them are.  They  might have some intriguing quirks, but their brains are essentially intact.  Yet no matter how much I plead, yell, cajole and demonstrate the behaviors I'm after, they don't quite get it.

Sure, the nicest thing one can do for any students regardless of species is to notice their best behavior and reward them for it.  The little girl in the first row of English class puts her name and the class period at the top of her paper for no apparent reason, but you like it, so you praise her for it and insist that everyone follow suit (thus ensuring future bullying as she quickly becomes Nerd of the Year).  Your dog brings you your newspaper because he mistook it for a chew toy, and you ignore the motivation and praise the effort, ensuring that he will bring you everything lying around the front yard for the next two days.

Your horse performs a perfect "back away from me while I open the gate" and you pat him and tell him there is no better horse on the planet.  He may or may not know exactly what he did to make you happy, so be prepared for a whole flurry of behaviors, one of which might be "back away from me" but many will be on the order of "swat a fly lift a hoof snort twice blink poop" because your praise wasn't quick enough to target the desired behavior.

The difference isn't in intelligence level among species but in the types of behaviors and their complexity and our inability to cut to the bone of the situation by using task analysis appropriately.  Even the paper-delivery dog may not be sure whether it was the bringing of the paper or how he was holding his tail when he delivered it to you, but those are small differences.

If we really want to put forth the best possible effort to teach our horses what we want them to do, we should take a page from dog trainers and spend as much time as we can actually shaping the behavior we want.  That takes time, energy and focus.

As I discussed last week, horses, though there seems to be evidence to the contrary, do not actually understand human language.  We can teach them to connect our vocal cues ("stand!") with a specific behavior by physically putting him in a spot, speaking the command, then walking away.  If he follows, we can say "No" (I don't consider that a curse word), and put him back in the same spot.  If he stays put for even a second, reward him.  If he doesn't, repeat until you're both sick of the process or he's done the behavior to your satisfaction at least 100 times.

Lucky for me, Duke thinks wearing the bareback pad is a reward.
He'll stand there all day without fussing as long as he can continue
stylin' like a rock star.

Clicker training advocate that I am, I promise that you can cut the learning curve down considerably by the use of the clicker as a secondary reinforcer.  If he stands for a heartbeat, and if you have taught him the basic "This sound means a treat is coming next", he'll be doubly reinforced, once when he hears the click and anticipates his reward, and once when the treat enters his mouth.

Why not just pat him on the head and tell him he's good?  Well, that's fine, and he'll appreciate it, but not as much as food.  Food rewards win hands down every time. Anything more complex (like giving him turnout after he's performed eighteen various behaviors under saddle) results is a more vague response.  He's not clear about exactly why he's being rewarded, and you are going to have to be absolutely certain that the reward you've picked for him is actually part of his personal currency--the stuff he works for.

So catch him being good for sure.  But don't neglect the opportunity to explain to him exactly what you're after.  The process will be more fun and go faster if you're both speaking the same language.

No comments: