Luring, Targets, and Shaping, oh my!
When I first started training with food, I used luring. If I wanted a sit, I held food over the dog’s head and waited for the butt to fall – saying “sit”. Dog gets a cookie. If I wanted a front, I took that cookie and centered it in front of my body. Dog gets a cookie. For heeling, cookie was held in front of dog’s nose and at my left side. We then walked when I said “heel” (the dog, cookie, and myself) and at various points….dog gets a cookie. You get the idea. That is luring; dog follows the food into desired positions which are simultaneously named. Over time, the lure is faded and all that is left is a cue – either verbal or visual. It is no surprise that most dogs are more proficient at the hand signal associated with the lure than with the verbal command, since the focus of their attention was on the hand holding the treat (as opposed to what you might have been saying at the time).
After using the luring method for a while, I discovered two shortcomings. First, dogs don’t seem to do much thinking when their nose is following a cookie. I had the vague feeling I was teaching everything twice (once with a cookie present and once without). Second, I learned that luring with a toy is close to impossible, because dogs don’t glue their nose to a tug or a ball and follow it around. And as a big proponent of toy training, I considered that a pretty big issue.
As I learned more about shaping and clicker training, I became less of a “luring” trainer and more of a “structured shaper”. (I made up that word…no point in looking it up).
What is a ”structured shaper?” A trainer who sets up the environment so that the dog is extremely likely to find the behavior that leads to the click very quickly.
Here’s an example of “structured shaping” in heeling: When your dog is somewhere behind you in a small and boring training area, walk away and call to your puppy while looking at your feet (not back at the dog) and circling to the right. Wiggle fingers (or clap) on the left side of your body and orient your head slightly to the left. If you picked a sufficiently dull area, your puppy will come up to see you, simply because there is nothing better to do. As puppy comes up on your left side (they naturally come to where you are orienting), click and feed puppy in perfect heeling position. If you use a toy, click and reward in that same position. If your puppy isn’t much for coming when called, make sure they know you have cookies or toys on you. I “click” for showing up and I “feed” in perfect heel position. I turn slightly to the right when I feed in order to teach puppy to drive into heel position rather than just showing up. If you look back at my first days with Lyra, you can see how I do this on her videos. Within a week, you’ll see that she has a pretty darned good idea about where heel position is, at least on a circle to the right or in a straight line.
I call that “structured shaping” rather than “free shaping” because I set up the situation to make it happen. “Make it happen” is one of my favorite training phrases. It simply means to make the choice you want likely without the dog having to think about what it will do. If you consider what I’ve said for a moment, you’ll realize that there is an element of “luring” in structured shaping. (My walking away clapping causes the dog to show up in heel position – that is luring with my movement, voice and body position).
If that is my definition of “structured shaping”, then what is “free shaping”? In my mind, an example of free shaping is “100 things to do with a box”. In this game, a box sits on the floor, and the dog is clicked for doing different things with that box, until the dog is offering a wide range of behaviors relating to the box (normally you’ll see nose touches, paw touches, body touches, climbing on or in, barking, etc.). As a handler, you do not influence the dog’s choices; you simply click and reward the dog for offering a variety of behaviors. Most clicker trainers would say that “structured shaping” is simply a form of free shaping, but I choose to differentiate them because I believe they lead to different end results. Structured shaping teaches a dog to rely on you and to look for environmental cues about what is desired. Free shaping teaches a dog to run through a repertoire, only limited by the dog’s imagination. Eventually, cues are added by the handler to gain control of the free-shaped dog’s behavior.
So where does targeting come in to this discussion? I see it as halfway between a lure and structured shaping. Lyra is taught to follow my finger to get a cookie, which makes my finger the target. Using my finger, I can “lure” her with my finger…with no food in it….into any position I want. That is what I found myself doing this morning when I started teaching her some static positions (sit, down and stand). Eventually I will teach a pop up stand, but for now I simply want her to move between a sit, down and stand without letting her front feet shift position at all. I decided a finger target was a reasonable way to accomplish this. I will use targets for other things along the way, but this is my first use of a taget.
Where am I today? I'm a structured shaper who does very little luring. I've given serious thought to doing more true free shaping and I have opted away from this option, at least for now. I'll talk more about why in later posts.