This came up in another dog training forum I frequent (reddit.com/r/dogtraining). Thoughts?
http://www.clickertraining.com/node/101
Hidden Aversives: Are You Punishing Unconsciousl y?
By Karen Pryor on 01/01/2005
It's a new year. Time for good resolutions, right?
Let's resolve to stop punishing our dogs by accident.
"But I'm a positive trainer, I don't punish!" you say. Well, I think we sometimes do, and don't realize it.
I'm not talking about reprimand. That's a social act. A puppy gets too rough and Mama dog growls. A canine nose gets a bit too near the cheese and crackers on the coffee table, and you speak warningly. That's communication, and it's legal, in my opinion (unless it becomes ALL you do, and replaces actual teaching—and I do know pet owners and parents who lean that way).
I'm also not talking about naturally occurring aversives (an aversive is any event that one might find unpleasant.) Here's Kay Laurence on that topic: "Dogs get plenty of 'aversives' in every day life. My Gordon Setters regularly run into the door in misjudging the gap, stub their toes, fall off the bed, go the wrong side of a tree, etc. Hey, that's life. But it's not good teaching."
Jumping up
What I am talking about is aversives that even "positive" trainers use deliberately, to affect behavior. Here's a popular one. A trainer asked me, "Karen, how about when the dog jumps up? Are you saying it is not proper clicker training (+R) to turn your back (take away attention) and then click him for four on the floor? This is an aversive, right? I only want to be 100% + trainer and I use this aversive technique. Please elaborate."
Punishers, like reinforcers, are defined by the receiver, not the giver.
I have a video clip of a trainer doing just that. She's working with a big yellow shelter dog that jumps up a lot. Twice, the dog offers a sit, and she clicks and treats. The third time, the dog sits, but the trainer waits a bit longer, and the dog jumps up on her. She folds her arms and turns her back. As she does that, the dog cringes backward to the floor, as if it had been struck.
Was that "punishment"? To the trainer, no; she just briefly removed her attention and what's so bad about that? To the cowering dog, yes: that really hurt.
Punishers, like reinforcers, are defined by the receiver, not the giver. I am sure the trainer thought she was just communicating, not punishing. But in a dog's world, licking faces is a puppy's way of making adults be nice to them; so dogs jump to get near our faces. Turning your back is, therefore, negative punishment: removing something the dog very much wants. The dog's behavior of cringing tells us, from the dog's point of view, this was painful.
Now this delightful clicker training experience, in which the dog WAS learning how to be successful, has become a mix of good and bad. Learning, even with food involved, is no longer such a good thing. The trainer herself is no longer an unmixed blessing either. That's what the introduction of punishment does, whether you want it to or not: it makes the learner insecure. At ClickerExpo, Jesus Rosales-Ruiz and his students have been showing us new research from the University of North Texas that suggests it doesn't take much punishment, and it doesn't take long, for that unpredictable mix of good and bad to taint the dog's enthusiasm for attending to you, and for learning in general, for a very, very long time.
"Guiding" with the leash
Here's another popular punishment. At ClickerExpo I see it all the time. The dog is on leash, quietly standing or sitting next to its owner. The owner gets up and starts to walk away. These owners typically don't look at the dog, they don't speak to the dog, they just start off, and as they do so, they jerk the leash. That's how they tell the dog, "Come with me." Even if the dog is already moving with them, bang! That little leash pop. Why?!
In Orlando I finally got annoyed enough with this to make a ClickerExpo demonstration around it. I picked someone with a large, pleasant young dog, who was doing this leash jerk faithfully every time she sat down or stood up or moved. We got a chair for her to get in and out of. I told her when to move, and I clicked her for giving the dog a verbal cue before moving. I tried to foil that well-honed habitual yank by putting the leash in her "wrong" hand.
It was hard for her; she really thought she had to give that leash pop. It was easy for the dog. When I finally stopped the popping by taking the leash off altogether, the dog, almost prancing, eagerly glued itself to the owner's side while she stood, sat, stood, and walked. "Whew! What a relief," said the dog's happy face. Everyone laughed. Obviously, the dog had learned very well how to pace itself to the owner's moves, and was glad to do so. The yanks were completely gratuitous punishment.
Teaching aversives as requirements
Someone, I could see, had put a good deal of effort into making that owner a consistent punisher. "Probably you've been scolded for NOT popping the leash every time," I ventured, and indeed heads nodded all over the room.
So, aversives are used on the owners, to make them use aversives on the dog. And does it do any harm? Oh yes indeed. In the course of one morning at ClickerExpo, these well-behaved dogs might receive dozens of unexpected aversives. Pop pop pop. They get a yank while lying down quietly (surely a good thing to be doing?) or while standing still next to their owners (another good thing) or while walking with them as they change direction or stop at the elevator door or start down a hall. There is nothing they can do to avoid the pop; even skillfully anticipating their owners' every move does not bring relief. No wonder they wear facial expressions of patient resignation, like most horses.
Major vs. minor?
The question is NOT how major or minor the aversive is. The question is, why use it when you don't have to! I quote Kay Laurence again (©2003 Learning About Dogs, reprinted here with permission): "If I am teaching a dog, I avoid every atom of punishment or removal of something good to get the behaviour. It is not a question of how aversive, it is the thought that aversive is a method to get a behaviour. The actions are an indication of the thought process that aversives are part of the teaching process. I will say, 'Let's just find another way.'"
We all think we're "positive" trainers. But training with reinforcement involves more than just being nice, and more than using reinforcers. It involves creating a climate of security in which it is safe to learn new things, and safe to rely on what you've already learned. In this climate an animal can learn to control itself, rather than being controlled by you. In this climate, rather than just reacting to the environment like an untutored shelter dog, barking at every noise, plunging towards every attraction, jumping on everyone and everything, mouthing and smelling and grabbing—an animal becomes confident and calm. In this climate, having confidence that your cues are meaningful and will lead toward pleasant goals, the dog is trusting and—this is very unscientific—the dog is happy.
A happy New Year for dogs
So, in this New Year, let's resolve:
1. Whenever an aversive seems like the "answer" to a "problem," find another way.
2. Give up the leash pop. Period. Don't use it as a tool, as a cue, as a correction, or as a habit. Speak to the dog, before you move. Then teach the dog that going with you, whenever and wherever you go, is a clickable behavior. It won't take five minutes. And your dog will be so pleased."