The opinions expressed in previous entries may or may not express the current opinion of the author.

Sunday, October 9, 2011

Positive Reinforcement

My husband is reading a book called Dog Sense, by John Bradshaw. He shares interesting snippets as he reads along, but the main gist of the book is as follows: We have all been sold a bill of goods with this idea of the dog as a sort of domesticated wolf, whose social life is rigidly controlled by a dominance hierarchy, and who will feel most comfortable if you "show him who's boss." Firstly, the latest research shows that wolves aren't pack animals who live within a military-style pecking order - they live in family groups, where trust and cooperation and flexibility are the norm. Secondly, dogs aren't really at all like wolves anyway. Thirdly, forget dominating your dog - positive reinforcement is where it's at.

Our first exposure to serious dog training was with a Koehler method trainer in Hyde Park in Chicago. All the neighborhood dog owners took this guy's class, and let me tell you we had a neighborhood of dogs who behaved with Teutonic precision and orderliness. Everyone came when called, everyone could heel off leash, everyone sat at the kerb. Oh yes, we were an impressive bunch, and for me the Koehler method has always been the gold standard of dog training. As Koehler himself said - reliability off leash is the standard by which you should judge training methods.

There was only one naysayer in our neighborhood - a Hungarian professor, whose family had been wiped out by the Nazis. He just couldn't stand all this barking of orders and rigid discipline. He was known to dive in to the middle of a training session, scoop the trainee up into his arms and commiserate loudly.

Now it looks like Hungarian Guy was right, as along comes the "latest research," and we find that all this yanking and yelling and pinching and showing who's boss is just not an effective way to train your four-legged friends. Sounds familiar.

Anyway, my husband and I have been on a campaign of playing Mr. and Mrs. Nice Guy to the dogs. And we've seen quite impressive results. Malcolm likes to run after trucks, and he's fond of heckling the horses. Since I've started lavishing praise and sweet talk upon him for everything he does, there's been a big improvement in his response if I call his name when he's off yelling at a horse or vehicle. You can see him trying to withstand  the magnetic force of my voice; but he can't resist it, and he turns and comes over to me. My husband has also noticed that the dogs come over to elicit affection more often - which is especially noticeable with Malcolm, as he is not a particularly outgoing dog.

Now, I can't compare this to a full-blown course of Koehler classes, which I'm sure would yield much more impressive results. However, the off-leash part of the training culminates in using a "light line" to fool the dog into thinking there's no leash attached when in fact there is. The dog disobeys - yikes! - and (barring Hungarian intervention) wrath descends. But the positive reinforcement also seems to set in place a force of action-at-a-distance. You can see it reeling the dog in, and it never gets tangled or caught on bushes.

Good Dogs
So ... I had occasion to think of all this today when I was riding Rose. Which I was doing because my husband and daughter started to complain that we have NO horses who are dead broke (as the unfortunate expression goes). Plus also, I just really had an urge to get up on a horse. Rose was very sweet about getting tacked up and seemed to enjoy the process. Plus also she's a lamb about mounting.

Once up, however, I realized that what we do in fact have is GREEN. This is only the second (I think) time I've ridden her, and the first time was with a halter and two thick, clumsy ropes for reins. This time I had the bitless and was prepared to be a little more fine-tuned.

I rode her in the yard with the other horses milling about - usually at a distance, although they did come over and interfere a couple of times. Rose was quite willing to listen, very blocked in the shoulders, and got a little irritated sometimes when I wanted her to do something she hadn't planned on. She's used to just riding out with George, which involves very little in the way of steering or direction on the part of her rider.

Whatever it was or is that people are supposed to do when they're training horses or whatever it was that I used to do, well all of that is kind of out the window for me these days. I started out not knowing exactly how we were going to proceed.

The positive reinforcement thing came into play, when I realized that the little tap on the horse behind the girth was no longer a resource I could fall back on. A few times Rose refused to move forward. And became a little cross at the idea, as it involved moving away from the other horses. So I just asked her to wait, pointed in the right direction. And eventually she moved forward, and we didn't come to an argument, which is the main thing.

I realized she's very blocked in the shoulders and that we could profitably have a very short ride, focussing on that issue. She's hard to "steer", and I discovered that rather than going straight to working on the bending while moving, it was helpful to come to a halt and draw her attention to her withers, which seemed to be all out of synch with her head and back. Once she felt herself a little more connected at the halt, she was able to move off into a more comfortable walk.

We worked for maybe 10 or 15 minutes, and then I let her go again. I think she sort of liked it and that she is so very clueless that any amount of work we do together will make a big difference. Right now, the idea that I'm communicating with her and making helpful suggestions is a totally novel concept to her. Hopefully as she becomes more accustomed to working, she'll start out with the expectation of a conversation.

The bitless does feel like rather a blunt instrument, but on the other hand, I can afford to be hamfisted and blunt, as I can't cause nearly as much discomfort as I could with a bit. Once Rose is tuned in more, I think it'll be possible to be quite delicate with the bitless bridle. The only other problem with the bitless is that you have to have a noseband, and I can see that, even loose, it constrains the movement of the jaw.

After untacking Rose, I left the saddle hanging on the gate. Bridget came over and knocked at it with her nose. So I put it on her, and then realized - hey presto! - I'd saddled her at liberty, which was something I'd been berating myself for not doing yesterday. It fell off again before I got the girth on, and I didn't bother putting it back, but Bridget and I worked (with treats) on backing, turning, and standing still while I move away from her head.

Then it was George's turn, as my daughter got him to take the foreign exchange student for a ride. Or I should say tried to get him, as he wouldn't be caught - most unlike him. She finally asked for my help, and I went over to him with my arms spread wide, halter dangling in one hand, saying, "George! This is your lucky day! You get to take the girls out!" He came right up and dropped his nose for the halter. Go figure.

He then adopted a rather resigned, martyred, but gentlemanly demeanor, and my daughter got the foreign exchange student situated on his back, and off they went for a nice long walk, nice for the girls at least.

And later, in the beautiful golden evening, I went out to visit with all the horses.


  1. May I suggest "Don't Shoot the Dog" by Karen Pryor? I thought that book was excellent.

    Food for thought: When applied tactfully the whip works exactly like the leg- they are both examples of negative reinforcement (when you get the right answer the stimulus is removed). Clicker training is positive reinforcement because you add a stimulus -typically food- after the desired answer has been given.

  2. I read "Don't Shoot the Dog" many years ago - I remember liking it a lot, and I seem to remember trying to apply a lot of the ideas to my children!

    I know you're right that the whip can be used tactfully and without force. However, the way I'm working at the moment, most of the time I'm not saying "Move forward on command!" Instead I'm saying, "I'd like you to move forward now if that's convenient for you." If the answer is "Not right now," then that's ok. If we're being chased by a bear, then I hope I'll manage to summon up enough convincing vehemence!