Hand targeting is usually suggested as a great behavior for new clicker trainers since it is easy to get and easy to define criteria for.
I guess I didn’t read the brochure carefully enough because I messed up hand targeting for one of my dogs six ways from Sunday!
From time to time I share in the blog mistakes I have made in the past, so I recognized a great opportunity here. My way of making lemonade.
Touching its nose to the trainer’s hand or an object is often one of the very first things beginning clicker trainers are encouraged to teach their dog. Here are some of the reasons:
- Usually if the trainer has a background in traditional training, this is a new behavior, and doesn’t carry the history or baggage accumulated with sit, down, and stay.
- It is very easy to get dogs to make that first step towards nose targeting, since they will naturally sniff things you present to them.
- The criterion of “hand touches object” is fairly easy to define, as criteria go.
- It can be extremely useful. You can use it to build a foundation for distance behaviors: go touch the target. Stationing. You can use it for “push” behaviors: closing drawers. You can use it to build excitement. And it’s the first of many useful body part targets, like foot, chin, shoulder, and hip.
- Dogs and people usually get a kick out of it, and it can be a relationship builder.
How could I wreck targeting? Let me count the ways.
I don’t know about you all, but sometimes I just don’t realize how messed up I’ve let something get until something comes along to highlight it. Since I train Sue Ailsby’s Training Levels, and they are so fabulous at building in generalization, it’s often a step of the Levels that highlights my booboo.
A Story of Many Mishaps
Summer is my crossover dog, meaning that I taught her in traditional ways that included aversives (e.g. leash pops, forcing butt down into a sit, prong collar) before switching to more humane and considered methods. So she has been through a bit with me, and has stuck by me so sweetly.
Summer first learned to target a “touch stick” as we called it in agility class. I made my own by painting the end of a wooden spoon navy blue with non-toxic paint. She thought it was fun and even back then I was able to send her 20-30 feet out to touch the stick, mounted in the ground.
It was only after a lot of practice with the stick that I taught her to touch my hand. This was all before I discovered the Levels. So back then I taught a touch right in the center of the palm of my hand (Sue encourages using the ends of the fingers).
I’m not sure when I noticed that Summer’s hand touch was not nearly as enthusiastic as her touch to the stick. Nowadays I theorize that it is part of her personal space bubble. During most training she would really rather not touch or be touched. She started giving me a feather-light hand touch, even verging on a drive-by. I probably shrugged mentally (if I noticed at all) and reinforced it. I usually used the hand touch to keep her engaged and I didn’t consider a functional purpose beyond that. (And didn’t think of the irony of the tentativeness of her “engagement.”)
So that was Problem #1: A huge reinforcement history for non-existent touches. Check.
So thennnn, when I first started doing the Levels, I really wanted to work on Retrieve. Summer would happily do the targeting behaviors in the old Levels (and I brushed on by our little problems with criteria), so I went ahead to work on Retrieve. I used a wooden dowel. Even though I held it far differently (and it didn’t have a blue spoon on the end of it), the dowel elicited her more hearty nose touch back. I was glad, and the beginning of retrieve starts with a nose touch anyway, so I reinforced it quite a bit.
Then for many months I tried to get her to open her mouth. With the help of my friend Lynn, I finally did get teeth touches and finally a slightly open mouth, but it was tough going for this novice handler with an “un-mouthy” dog. Then I stuck with those teeth touches too long and gave them a huge reinforcement history. When Summer would start with a nose touch, I would wait for what she tried next. Usually it was a touch with a slightly open mouth.
So Problem #2: Big reinforcement history for teeth touches that were never put on a cue. Check.
Around the same time, I started working on foot targeting, also in the old Levels. Summer LOVED whacking things with a front paw. We included objects on the ground (for example the Easy Button), and marks or tape on the wall. This at least I did put on verbal cue but I’m afraid most of the time she was cuing situationally. You can see her very enthusiastic paw touch in this video, and also check out my really lousy timing. But at least I was late rather than early. (Early can contribute to drive-bys or other tentative or incomplete behaviors.)
So Problem #3: Practicing lots of the fun paw touch behavior–**WHACK**–while the nose touch was drifting into a drive-by. And the paw touch did not have a fully learned verbal cue. Check.
Then I took Susan Garrett’s Recaller’s class. She emphatically discourages using a verbal cue for hand targeting, since she observes that in that case often her agility students will tend to cue it to get the dog’s attention when it is wandering. This can possibly give tertiary reinforcement to the dog for not paying attention. She wanted her students to use it only when the dog was already paying enough attention that it could get very engaged. So I stopped using the verbal cue for a while.
This was Problem #4: Being inconsistent about the verbal cue for nose to hand targeting. Check.
I also worked hard on a puppy fetch behavior with Summer early on. I really wanted her to bring me things. I knew this wasn’t a formal retrieve and I didn’t approach it that way. But in the course of teaching Summer to bring me a little basket, after I finally got her to open her mouth and take it, she had thousands of “fetch” reps for high value treats.
Problem #5: Big reinforcement history for biting and carrying things. Check. Again, this is only a problem because I had let the nose touch stay so weak and let the verbal cue slide.
Finally, I’ve written before about Summer’s huge reinforcement history for looking at the treats or my treat hand. I realized while recording video for this post that she is not looking at the target most of the time at all! She is just following her nose.
Problem #6: She’s not looking! And that has a big reinforcement history as well! Check!
Analyzing the Errors
- Timing has to do with how accurately and consistently in time you mark what you want, and also how quick and accurate your treat delivery is.
- Criteria has to do with how well you define the behavior for the dog, and hold to that definition.
- Rate of reinforcement has to do with reinforcing with a great enough frequency that the dog is eager to participate and learns the behavior solidly, but it also has to do with the rate compared with that of other behaviors.
Just for fun (I know, I’ve been told I have a very weird sense of fun) I made a chart of my failures. X’s mark the errors.
The Drive-bys were partially a result of timing (clicking early), and largely a result of not holding firmly to the criterion of nose-touches-hand. Teeth Touches were the result of settling for too long on an interim criterion, and were too heavily reinforced. The Paw Touches and Biting/Carrying were good behaviors, but were reinforced much more heavily than the nose touch. The inconsistency of dropping the verbal cue for a time was a criterion issue. And Not Looking–that was a criterion problem again, and an unwise thing for me to reinforce.
So hey, do I win something? Most mistakes with one simple behavior?
My friend and thoughtful trainer Mary Hunter has a wonderful post on targeting called, “The Continuous Nature of Targeting.” She describes the smaller elements that comprise the nose target behavior, and how she deliberately experimented with the timing of the click to help a horse with some anxiety issues.
A more experienced trainer than I wouldn’t have the problems with the other reinforced behaviors that I did. They would probably have gotten those behaviors on cue, and a better trainer wouldn’t have gotten lax on the nose target itself. The result of the combination of errors I made was that if Summer does a light nose touch and I hold out for a better one, she defaults to offering those other heavily reinforced behaviors. With many dogs, withholding reinforcement in that situation would result in a harder hand touch to follow. We often count on that response in training. What Sue Ailsby calls the “Hey, stupid!” reaction from the dog. However, because of the other reinforcement histories, with Summer instead I will usually get teeth or a paw instead. And if I don’t reinforce that, I will often get a flurry of chaos. You can see a little of that in the movie.
Also in the movie you’ll see that paw movement has actually become a superstitious part of her nose touch. It is there much of the time. But in the course of recording footage for the movie, actually Summer’s targeting improved! I guess putting myself on camera made me do better with criteria. So while I still have some rehabbing to do, it’s not as bad as I thought. For instance I found out that she has learned that when a Teeth Touch doesn’t work, she can close her mouth and do a nose touch.
For my retraining plan I will choose positions that are not conducive to biting or pawing. I’ll make it as easy as I can for Summer to target with her nose and make it more difficult to do the undesired behaviors. She tends to mix those in when I use a stick or something on the wall, and not so much with my hand, so we’ll start back at the beginning. (The order of things to touch in the Levels is hand, foot (the human’s), wooden, plastic, and metal objects, a sticky note on the wall, then a cabinet door.)
I’ll get the treats off my person at least part of the time (unfortunately I probably can’t use the Treat and Train with her). And when I get to the “touch a spot on the wall” step, rather than using a sticky note or painter’s tape I’m going to paint a spot on some cardboard so she can’t grab it!
The Good News
You know what the best news for Summer is? That I was training with positive reinforcement. I reinforced way too big a variety of behaviors, true, but even so, I can fix it with minimal frustration to Summer. Being fair to her will be my first criterion for any training plan I make to address the last odds and ends of the wrong behaviors creeping in. Although it would be tempting to use negative punishment, for instance, pulling my hand away for a moment if she uses teeth or a paw, I would rather not if I don’t have to. It dampens her enthusiasm and just seems unfair since I’m the one who reinforced those things to begin with.
It’s interesting that Summer, a crossover dog, will offer all these weird behaviors. You can see her in the movie holding her cheek up to the wall when I have the tape on there. That would not be an easy behavior at all for me to train on purpose. But I got it by accident! She does a lot of stuff like that. I’ve spoken in the past about a “map of reinforcement.” You can see what behaviors have been reinforced in an animal. Summer’s is a very interesting map!
The Moral of the Story
Be careful with your foundation behaviors. It’s worth taking the time to get them just right the first time. If you are doing the Levels, do them in order. Don’t blow on to the next thing leaving little mistakes! They will grow.
But also I hope no one is daunted by my sharing my mistakes. My intent is the opposite. The point is that they can be fixed. Also, I am not that embarrassed by my shopping around and doing a little of this and a little of that when I first crossed over. The Kid in the Candy Shop reaction to force free training is a pretty common one, I think. It’s so cool when your dog starts offering behaviors!
And the up side is that Summer does have a palette of unique things that I have accidentally reinforced. One of these days I’ll take advantage of that. On purpose!
Does anybody want to compete with me on the most mistakes made with one behavior? Bring on the comments! Would love to hear about successes and problem solving, too.
Thanks to my friends for the photos:
- Marge Rogers of Rewarded Behavior Continues
- Sonya Bevan of Dog Charming
- Louise Kerr of Elite Pet Care and Education
Thanks for reading! Coming up:
- Shut Down Dogs (Part 2)
- Threshold: Let’s Work on Defining It
- Clara’s Rules
- OMG Could She Really be Talking about the Continuum AGAIN?