Sunday, May 4, 2014

Why We Reward Choice Instead of Using Force

There's a common thing that you see with some novice dog trainers and also a certain kind of experienced person who maybe hasn't read a dog training book in a decade or so. When they are teaching a dog to sit, or when a dog who theoretically knows sit but does not obey the command, they push down on her rear end.

As a teeny puppy, Jax learned to volunteer a sit with terrific focus.
First, let's remember that we're not actually teaching dogs to sit. Dogs are perfectly capable of sitting without the intervention of a professional; what we're teaching them is to reliably volunteer a sit in response to a cue. The cue in this case is the word "sit," a hand signal, or both. And in order to do that, we have to get the dog to understand that the cue means we want the behavior, and then we have to make it rewarding for the dog to execute the behavior so she wants to do it again in the future.

So if you give the cue and then physically force the dog into complying with it, it may look exactly like you've just gotten a sit, but since the dog hasn't volunteered it, you've taught her very little, no matter how gently you push her rear down. And indeed, since the natural reaction from a dog in this situation is to push back, you may not end up being all that gentle. Lastly, pushing down hard on a dog's hips, especially when she's young, may risk injury.

The fact is that forcing her into position not only fails to teach the dog to sit, it may also teach your dog something entirely different and entirely undesirable: that when you say "sit" you're going to grab her and push on her. This method has two common results: a stressed out dog who doesn't learn to sit at all but starts showing signs of stress, or a dog who sits because she's figured out that if she doesn't, the unpleasant sensation of being grabbed and pushed isn't far behind. You may end up with a dog who sits on command, but your likelihood of a dog who is happy and reliable about it are a lot lower than if you rewarded a choice.

You can teach a dog she has to obey because if she doesn't, you'll do something unpleasant, but that's not a great foundation for a strong bond with your dog, and it doesn't lead to truly reliable behavior. For example, I've noticed that dogs whose owners resort to force often have trouble using the skill at a distance. If your dog sits in order to avoid having you grab him or push on him, it can be very difficult to get that dog to volunteer or hold a sit once you're more than an arm's length away.

Phoebe was rescued by a gentle owner who focuses on rewarding good choices.
Plus, even though many dogs will become compliant when forced, some will start to develop anxiety around people and may even develop reactivity and aggression because they don't trust the humans around them. The less force you use, the lower the risk of reactivity, but why risk it at all?

This same principle applies to all kinds of skills. If you want your dog to learn to sit back up from a down, don't pull her up with the leash. You can lure with food, you can clap your hands, you can gesture, you can use your voice, you can get creative. If you subscribe to a concept called "free shaping" (more on that another day), you wait for your dog to spontaneously offer a behavior and then mark and reward it. There are all kinds of great ways to teach a dog to sit up from a down, but the common element in the successful ones is that you get the dog to voluntarily change her body position and then you reward the behavior.

By the same token, if you want your dog to walk by your side, steering her there with the leash doesn't teach her where you want her. It just teaches her that you'll pull her by the neck without giving clear instructions. The more successful, gentler methods involve rewarding the dog when she makes the choice to walk in the loose leash zone by your side.

In all of these examples, you're looking for an opportunity to reward a choice, which means the dog is more likely to make that choice again in the future. I can speak from experience that the handlers I see who reward a dog's good choices are the handlers who get faster results and more reliable behavior. And even with trainers who are somewhat old-fashioned in teaching some skills, the skills they get better results on are the skills they've taught and reinforced by catching the dog's good choice and rewarding it.

At a recent seminar with Ian Dunbar, this puppy worked off-leash in a room full 
with 43 other dogs. The owner's challenge was to create a rewarding connection 
with her pup using no force at all.
So the next time you find yourself pushing or pulling your dog into position in some way, put the leash on your own neck and give it a yank to remind yourself that you're not teaching fairly or productively.

When teaching and proofing behaviors, use your hands only for delivering rewards like petting, play, and treats, not for prodding, pulling, or poking. If you find yourself using the leash to steer, catch yourself and remember that you're only really teaching your dog if the tension is completely off and the leash makes a nice J shape. In fact, if you're in a safe environment, take off the leash entirely! That will put the onus on you to communicate with your dog using your body language, your voice, and your previous training. Just don't use an angry or stern voice: that's force too.

Dogs trained with low or no force methods tend to progress faster in basic skills like loose leash, sits, downs, stays, recall, etc., and they become more reliable because they learn how to work with their handlers to make good things happen. They also look happier doing it, and their handlers are happier because they're not choking or scolding their dogs. It's a lot more fun to train if you're essentially playing with your dog rather than punishing his failures—which are probably actually the human's failures anyway, but more on that another day.

No comments:

Post a Comment