Wednesday, October 31, 2007

Episode IV: A New Hope (part 1)

Forgive the post title, but I attended a Star Wars party over the weekend. And this is our fourth post on The Cat Incident and its fallout.

(With only a few minutes' notice that we were attending, I dressed as Callista-Sue Jade. Photo upon request. If you don't know who Callista, Mara Jade, and Mary Sue are, be glad of your social life. ;) )

Back to Laev.

Sunday afternoon, Laev spotted a cat outside the fence and began running the fence line, losing all semblance of control. I know that dragging her away from the cat only creates a stronger desire, born of frustration -- hence the adage of, "Always leave 'em wanting more!" And it's not that simple to catch a Laev in full cat-mode, either. I read somewhere that tests showed the aural-processing bits could actually shut down during true frenzy, and I think I have a sample case.

So I thought I'd try extinction. Let's see how long it takes for Laev to give up on the cat she can't reach, right? Couldn't be too long? The cat was sitting quite still, just waiting about 30' outside the fence for a human to happen by and feed it supper, so she wasn't actively stimulating the dog.

After about 30 minutes I gave up and went inside. After over 3 hours had passed, I didn't want to wait any more. I gave up on the idea of getting a baseline for extinction ("3.5 hours +") and went back outside with a clicker and a can of dispensing cream cheese. (The cat, by the way, hadn't moved; she was sitting exactly as I'd left her. So Laev had kept it up that long on just the visual, not any movement.)

Sometimes I wonder why we even bother to train for the AD*, when it's obviously so natural and easy for a fit dog.

I stood somewhat behind Laev. (I'm not foolish enough to risk my knees by standing in the path of a prey-frenzied dog; I've known two trainers who spent months on crutches after similar encounters and another pet owner who needed months of recovery.) I waited. I waited. I said nothing. Laev slowed to a long trot instead of a gallop. I waited. And then Laev paused, four feet on the ground at once, and I clicked.

No response from her, but I didn't really expect one, either. She continued running the fence. This went on for a while, my clicking whenever she happened to pause for a half-second or so and then her running on.

At last, though, she tipped her head toward me as she passed and I clicked and held out the cream cheese and she STOPPED. I quickly clicked and treated repeatedly, squirting cream cheese into her mouth as rapidly as I could click. Then I hesitated, she glanced up, and I clicked for eye contact.

Ah! The beginnings of communication!

She couldn't hold herself still for long, though, and she went off to run the fence again. But this stretch of running wasn't nearly so long, and soon she was able to pause by me again. I was quite pleased.

I gradually raised the criteria, requiring eye contact for clicks and then sustained eye contact, up to two seconds. Yes, that sounds tiny in print, but trust me, it was big! She kept glancing at me and then hurriedly looking back at the still cat. (The cat was fed by now, eating happily in place.)

I asked for a sit, and that was too much -- she hesitated and then bolted. Conflict must be resolved in movement, with Laev, and if she couldn't move from a stand to a sit, she had to run.

I waited (what had I to lose?) and clicked her back to focus again. This time she could sit, and follow me a step or two. Everything was fine 'til the cat moved.

Laev went berserk, screaming and biting the wire of the fence, trying to tear it out of her path. (This after *hours* of running?!) I waited; it's not as if I could do much to interrupt that, anyway. It would take an awfully strong aversive, and I don't want to go there. Fortunately the cat was content with the chunks of solid cheese I threw across the fence and she settled down about 15-20' away and watched with feline amusement.

I gradually clicked Laev back to focus and rebuilt our short spans of attention and movement. I called her to front, parallel to the cat. Cream cheese reward. I turned after a couple of reps and asked her to face the other way, still parallel to the cat. No problem. Then I took a giant leap and stepped back from the fence, asking her to front with her back to the cat.

It was almost possible to see her brain smoke, but she did it. She kept glancing over her shoulder, but I kept the cream cheese coming as fast as possible in short spurts, and she stayed with me. We went back to parallel tasks (sit and front), with heavy reinforcement for looking at me instead of the cat.

Our crowning achievement was working up to two steps of heeling, with full eye contact, with Laev between the cat and me (looking away from the cat to me). I reinforced massively and then led Laev, step by step, clicking and treating at nearly each step, away from the fence back to the house -- which is at least a couple hundred feet and that's a lot of clicked steps, if you ever wondered.

I wish desperately I had video of that session. We achieved a lot, and I'd love others' feedback on it and suggestions for improvement. Also, it'd be nice to have proof of what Laev's prey frenzy is really like, as it's hard to describe, so I don't always get relevant advice.

But Laev makes me -- forces me to be -- a better trainer. :)

* The AD is an endurance test of physical fitness and ability, consisting primarily of a 12.5 run alongside the biking handler.

1 comment:

Anonymous said...

Oh my! Laev sounds exactly like Yipper, though Yipper is probably a bit less persistent. How long has it taken you to get this far in YOUR understanding of this type of training....your mastery of it seems quite impressive!
Keep posting...I'll keep reading!
Great blog!
Jesse Ranee & Yipper