Category: Getting Started, Puppy Stuff, Performance Events ¤ Author: Dani Weinberg ¤ Title: Loose Leash Walking ¤ Carmel posted a message last week about how people using/adapting Dawn Jecs' Choose to Heel method. I've been thinking a lot about this. Our German Shepherd Ruby is now close to 5 months old, and she's been my "experimental" first all-clicker puppy. It's been a wonderful learning experience to start a new puppy and test out the things I teach to my beginning students! I've developed a really nice way of teaching loose-leash walking for my students, using some of Dawn's CTH, some Ian Dunbar (is quite complementary), and some I-don't-know-who (maybe Dani Weinberg?). It's really working well. But Ruby is going to be a competition obedience dog, so I'm doing things rather differently with her. Mostly, I've been experimenting - no, that's not the right word, because it implies a scientific and methodical approach, which is not at all what I've been doing. To summarize what I've been doing, let me call it a combination of Dawn Jecs, Karen Pryor, and Patty Ruzzo. I started her Dawn's Follow Food and Find Target at about 12 weeks but was not happy with it. Ruby was too short, and I was having to distort my position to offer my hand as a target. So I stopped for a while and picked up again when she was 14 weeks old. This time I went directly to Find (targeting to my left hand, with no food in it). Pretty soon, I stopped my hand as a target because 1) she didn't need it and was finding anyway, and 2) she was leaping at my hand - something I certainly didn't want to reinforce (and probably still a function of her short stature). I used the target hand only to guide her through About Turns. (I've been teaching hand targeting separately and plan to use it, others, to move her around the ring and set her up for the next - something like Dawn's "Get It" exercise.) I also used a wall on her left side to make sure she was really in good position when I reinforced. Somehow, she herself the idea of giving me eye contact while walking and of doing Sits occasionally. So now my criteria were: be on my left side with your head and shoulders in Heel position (I couldn't really see the rest of her was doing but decided to have faith!) and give me eye contact. >From that point on, I was using Karen Pryor's simple OC method of reinforcing for criteria and continuing to walk (faster and, if necessary, changing direction a la Ian Dunbar) if she was not in position. Several things started happening. When she was in position, she was wide and I let that go for a while, deciding not to introduce that criterion yet because I was most concerned with her understanding position with eye contact. When I stopped to c/t, she would pop out of position (unless there was a wall on her left side) and I'd lose her, so I would do an about turn as a way restart walking. When Ruby was almost 4 months old, I had an opportunity to work with a friend. We did CTH with distractions, and Ruby got it very quickly. Within just 2 or 3 tries, she was essentially ignoring the distractions and staying with me. Remember that this was after a long interruption in doing real, full-blown CTH with her. I also realized that Ruby didn't understand another new criterion that I myself was unaware of until I noticed her behavior: I want us to start together. I want her to watch me for a body cue (I'm not using any other cues up to this point and may decide to use only body cues) that would tell her that we're about to walk again. So I've been working on that part separately by using her eye-contact cue "Ready" when I'm quick enough to catch her sitting in Heel position. She gets her c/t for that behavior, and then, while I have eye contact, I take a step and c/t that, and so on, adding a few steps at a time. I've gotten up to about 5 steps so far, working just this tiny piece of heeling. And this is, of course, very conventional clicker training. Two weeks ago, when Ruby was 4 1/2 months old, we were training on a tennis court and suddenly we started heeling - I mean really heeling! We were moving in concert, she was in perfect position (no longer wide), giving me eye contact and doing a high-stepping German Shepherd flying trot! I was absolutely amazed and thrilled! During this past week, I've had a couple of opportunities to do some reality testing and have found that this was not a fluke. I even have it on videotape! When it first happened, on the tennis court, I stopped to c/t after only about 8 steps, not wanting to risk her missing her R+ for something so wonderful. Since then, I've been watching to see how long she can go, and it seems to be somewhere around 15 steps - though, once again, I haven't been pushing it beyond the "safe" point. Now is the time for me to start being uncharacteristically systematic and introduce VSR! I don't know how this happened. Could be, it's just an accumulation of both apparent and latent learning with the pieces coming together. Some research on learning has shown that a lot of learning happens underground, so to speak, like the bulk of an iceberg, and then seems to surface suddenly in an "Ah Ha!" experience. But that may be just an artifact of perception. Another possibility that I don't know how to explain in terms of learning theory or OC is that a complex behavior like heeling must, at some point, happen in an integrated way, all at once. I had a similar experience learning to do the Crawl (or Freestyle, or whatever you call that swimming stroke). As a child, I had been taught all the separate parts of the behavior - the kick, the breathing, the arm strokes - but it wasn't until I took swimming lessons again as an adult and my teacher said, in effect, "Just do it," that I really learned to do the crawl. Swimming and heeling are not really behavior chains but, instead, complex behaviors consisting of several parts that must occur simultaneously. Carmel, you talk about Dawn's Food Delay: >>That's what we've been doing and it works like a charm. I now have introduced this "IN THE BAG" yell to the dogs in other exercises too. Like the retrieve. And this keeps them very alert to see whether the treat at the end of the exercise will come from the pocket or whether there will be a dash to the BAG. By the way, I have found this a useful part of warming up because you and the dog can really push yourselves and run fast to get to that bag.>> I haven't yet tried this, but I will. Part of me is very reluctant to create such a huge interruption at this stage of the game, but you may be right about the value of varying the delivery of reinforcement. >>I am still having problems with the idea of rhythm heeling and DJ's reference to feeding as you heel for these long stretches. I think this raises the old debate about whether the click marks the end of the behaviour "only" or whether it may also sometimes be a "keep going" signal. What are other people doing with the really long heel sessions - stop mid session for c/t, or click and feed on the wing, as it were?>> I consider Rhythm to be a rather advanced step in the process. I use it as a pre-ring warmup with Sweetie, my older GSD who is training in Utility. I find that it's very much a self-reinforcing activity for both of us (like you, I move fast), and I do stop occasionally and interrupt with a c/t. I'll have to go back and reread Dawn on the subject to see if this is just my own interpretation. >>Enough for starters!>> Probably too much in this post of mine! If you've read thisfar - thank you! And thanks, Carmel, for raising the subject - one that I'm very interested in. Dani Dani Weinberg Albuquerque, NM and Crested Butte, CO