Lesson 9: (Control) Premack Principle Overview and Training Ideas

[s2If current_user_cannot(access_s2member_ccap_course_001_bm)]

To view this lesson, please purchase this course or log in if you have already purchased it.[/s2If]

[s2If current_user_can(access_s2member_ccap_course_001_bm)]

PREMACK PRINCIPLE OVERVIEW

premackDavid Premack was a psychologist and primate researcher who discovered a new kind of reinforcement in, shared in 1959. This principle is called Premack’s principle but is usually referred to as “the Premack Principle” in dog training. It’s also called the “relativity theory of reinforcement,” which I actually really like as a name. Everything is relative.

Here’s a small version of David Premack’s principle (“behavior” can replace “activity”).

High probability activities reinforce low probability activities

Or as I like to just say: “permission is a reinforcer.”

It turns out that having to do an actiivity that’s lower probability will punish the first behavior as well (make it less likely in the future). Like if every time you were on Facebook, you had to do your taxes, you’d probably be on Facebook a lot less.

There are other pieces to the principle but the bit about reinforcement is the part that we use the most in training. Premack’s theory is simple and yet very powerful, when put to use. Left to his own devices, what would be your dog’s most likely activities? Those are the high probability activities.

Access to those activities reinforces any behavior that’s not so likely in an untrained dog: waiting at the door, walking on a loose leash, sitting on cue, etc.

So any time you want to reinforce without food, think about Premack’s principle. If you can let your dog do an activity that they already do with greater frequency than the one you are trying to train, you’re in business. Grandma was using Premack’s principle when she said, “eat your peas and I’ll let you have dessert.”

We’ll talk more about this principle in the next lesson, but I want you thinking about the possibilities this week.

TRAINING: HAND TARGETING

Touch: With the Touch cue, your dog touches her nose to your hand. Touch is useful for getting your dog’s focus, for moving her from one place to another, for getting past distractions, and even getting your dog to come to you. Think about it. In order for your dog to touch her nose to your hand, she has to turn away from what she is doing and bring her nose to your hand. This automatically brings the rest of her body with her. It’s possibly THE most useful skill for a dog to have.

To teach Touch, start out with a treat and the clicker in one hand. You can have your target hand a flat hand or whatever you want, but I like to use two fingers as a target. Make a fist with your other hand, except put your pointer finger and middle finger out. That will be your visual cue for Touch. Present that hand to your dog. Your dog will then probably go toward your hand, expecting a treat. Ignore any pawing.

When she touches the hand with her nose, click and treat. While she’s eating, put your hand behind your back and then present it again when you’re ready to click and treat again. For some reason, it makes your hand brand-new and interesting again.

Do this several times before saying a cue. Once she’s got the hang of it, and you’re relatively sure she will touch your hand, start saying Touch right before you put your hand out. Continue to click and treat for touches.

If your dog stares at you and doesn’t touch the hand, then either wait her out or put your hand behind your back and bring it back out again. Don’t lean into her or stare (that’s a bit scary). Your hand may also look like a hand signal you’ve already been giving her. If that’s the case, change this to a new signal—hand flat, only one finger, etc. If she is biting your hand rather than gently touching with her nose, make sure you aren’t clicking for the bite. Click sooner to reward her before her mouth opens, or click later, waiting for her mouth to close before clicking.

Begin to move the target a bit, so your dog has to walk a step or two to touch your hand.

Your exercise this week is to alternate between Touch and Find It. So when you click for your dog touching your hand, say Find It and toss a treat on the floor for your dog to find. Start with an easy location and build up to tossing it several dog-lengths away. I do this for two reasons:

  1. It moves the dog away and sets you up for another Touch, automatically building distance into the Touch cue.
  2. Find It is probably my second favorite cue for dogs to know. You can use it as a distraction, to reduce stress, and for physical and mental exercise.

Note: if your dog is not familiar with Touch, keep quiet and don’t say the Touch cue until the behavior is reliable. You can, however, immediately use the word Find It, because that is already a highly likely behavior and you can make it happen by walking near the treat or tapping it with your foot if your dog really isn’t getting it. Try not to do the dog’s work for her…let her find it!

If your dog already is great at the Touch cue, increase the difficulty level by practicing around a bigger distraction. After 5 marked/treated Touches, mark the last one and release your dog go check out the distraction, using your All Done cue (and then maybe walking or pointing toward the distraction, if needed). Pop Quiz: what principle are we using here?

TRAINING: MARK AND MOVE AND OTHER SURVIVAL SKILLS

Mark and Move is covered in detail in BAT 101, the prerequisite for this class. It is also discussed in the free Survival Skills handout on public handouts page and my Ahimsa Dog Training Manual.

In short, you basically mark some behavior (click, yes, done?, etc.), move away, and give a reinforcer.

Mark and Move has an added benefit for dogs with frustration issues who want to run over to other dogs, for example. It reinforces the behavior that was marked, but also reinforces moving away from the other dogs, because the reinforcer comes AFTER the movement away.

[/s2If]