We are just finishing up a month of looking at the terms we use in training. I’m a bit late with this but better lat than never.
I thought it would be fun to take a look at the words we’ve been studying in action. See how they apply in practice. And maybe some words we didn’t talk about but that we should know.
This video starts out with negative punishment. When we hear the word punishment we think bad. Not something we want to use! In scientific terms is simply means something that stops a behavior from happening. In the same way that a reinforcer encourages a behavior to happen again. The negative here doesn’t mean bad, it means removing something.
In short negative punishment means stopping a behavior from happening by taking something away.
I don’t want Harvey to pick up the cone so I am not rewarding him for it, I’m removing the treat he hopes to get.
Then we try again. I click too late. I was trying to click before he bit the cone. Instead I end up rewarding what I don’t want. Yes, this messes up our training but once we’ve clicked the horse gets a reward. Those are the terms of our bargain with the horses. I will not break them, I’ll just do better next time. I can just give one tiny pellet though, a tiny reward.
Then finally we both get it right! I click before he can pick up the cone and reward enthusiastically.
A bit of treat manners in there. Never ask for a trick while they are mugging you! The trick is a tertiary reinforcer, something that tells them they will soon be getting the secondary reinforcer, the click, then the reinforcer itself, the treat. That’s a lot of different names for a reward. We can break that down into simpler terms. The trick tells them the reward is coming, making the trick itself a bit of a reward. If we ask for the trick while the horse is mugging us we are rewarding the mugging and teaching bad treat manners.
Then Harvey gets the behavior I want right again! We never want to stop too long at one part of the whole picture. As a general rule getting the behavior right three times in a row tells us it is time to move on. Harvey got it twice. I know my horse and am happy to come back to this as often as we need to. I decided to throw in the next piece of the puzzle. I asked him to touch the card in my hand. I will chain the two behaviors together to make one trick that will only be rewarded after both pieces have been done.
To start with I’m going to reward for both, but only for a short time as he begins to understand what I’m asking. If I reward for both for too long he will get upset at only being rewarded for one. I will go back and forth for awhile rewarding after the last behavior and rewarding for both as we refine the exact behavior. I will need to reward for NOT picking up the cone quite a bit before he can just look at the cone then choose the number it matches.
Harvey is such a quick learner that he gets the idea immediately! He looks at the cone then chooses the right number after one try. That doesn’t mean we are done. This is a good place to stop though. Horses learn best with a bit of dwell time to think about what you worked on. It isn’t the end of the trick either. He needs more than one number to choose between for there to be any actual choice. The end goal here is for him to count the cones and choose the number that tells us how many cones there are. Lofty goals but by breaking things down and using all these tools at our disposal it is something he can easily accomplish!