Sunday, January 15, 2006

Communication

My last post gave me an idea. Training gives animals control over their environment (sorry about the anthromorphization, Dr. Machado). It gives them decision power - in a certain situation, they know that by doing something they will get a treat (or whatever). If they don't do it, they won't get the treat. It's really a form of communication. As I stated, Athena seems to need the communication and control, showing this by her insistance on routine and my opening doors at her demand.

I know it seems strange that by MY training HER, I'm giving her control - especially since the examples that I'm using go the other way. Trust me on this. She's not going to see it that way.

As Jonathan said, when I trained Travis, "What if he thinks that by [performing the action] he's training you to give him tuna." My response was, "He'd be right." Training by positive reinforcement isn't so much one animal forcing another animal to do something as it is a rough-and-ready form of communication. I say "sit," but what the cat is hearing is "If you want tuna, you can get some if you sit." Then they have a choice about whether they want tuna or not.

This also breaks us out of the cycle of our current relationship. So far, it's mostly composed of my giving her treats and trying to pet her and her asking me to open the basement door so that she can go downstairs and hide from me. I should have seperated her from her family as soon as she could eat solid food, but it's a little late now. I hereby do solemnly swear that I will not try to touch her at all during the training sessions. Thus, this will be a "safe" space for her to interact with me without her usual worries.

So - I'm going to "train" her to go to the top bed of her cat furniture on command. It's something that she does already, so it's a point of least resistance.

When I trained Travis, I started by having him jump up on the kitchen counters. Then I had him jump off. He could sit on command, and I was getting him to turn around when we stopped the training. This made it easy for him to understand
First, that he needed to do something for a treat.
Second, that he needed to do different behaviors at different signals.
Third, that when he saw a new signal he needed to watch me for cues on how to perform.
He was a really smart cat, and very trainable.

I think Athena is trainable too, but for different reasons. Travis's behavior was very variable, so he had figured out every possible way to get table food (and a few other things) when I started training him. For him, training was a new way to get table food. Athena is treat-motivated and likes rountine and rules.

I started Pavlovian-style clicker training today. Click-treat, Click-treat. It's impossible to get animals a treat in time to reinforce very specific behaviors, and if they come to you for the treat, they are associating coming to you more than the behavior. The clicker is a clear, fast, distinctive way of telling an animal "Yes." She left the last treat on the floor, so I've got about 6 - 8 chances in one training session. That will keep me from overworking her.

She gets it. Tomorrow, I'll do 3 clicker-treat pairings, then start asking her to do stuff for the treat/click. Let's say 4 training trials.

If I don't get good behavior response from treats, I'll see what tuna can do.

No comments: