05 MarEpisode 125 | In Defense of Negative Punishment: Teaching dogs patience at mealtime, stopping thumb sucking & more
Is all punishment necessarily bad? Nope! There are two kinds of punishment, and one of them is actually frequently used by so called "positive reinforcement" trainers like Annie: Negative Punishment. Here Annie breaks down what Negative Punishment is, talks about how it interplays with positive reinforcement, and explains its role in the "Elevator Game," which is a great exercise you can to do at mealtimes in order to teach a dog to not bum-rush the food bowl. She also reads from Behavior Principles In Every Day Life by John D. Baldwin and Janice I. Baldwin, which talks about a study in which parents used negative punishment at story time in order to discourage children from sucking their thumbs.
Subscribe: iTunes
Mentioned in this episode:
Behavior Principles In Every Day Life by John D. Baldwin and Janice I. Baldwin
Other episodes and blog posts on this topic:
Episode 65 | Don't Let Your Dog Cry It Out: On Training Dogs To Be Alone | School For The Dogs
Episode 47 | A busy person's guide to operant conditioning | School For The Dogs
The Big Bang Theory explains Operant and Classical Conditioning
Dog Training Lessons Learned From Watching Girls
Transcript:
Annie:
Hi, I wanted to let you know that I have a brand new, totally free masterclass available. And I'd love if you wanted to check it out. It's about an hour long and it goes over three simple things that every dog owner needs to know in order to teach a dog quickly and easily without pain, force, a major time investment or fancy equipment. When you register, you'll also get a free 20 page e-book all about what I call the dog training triad. You can find it at Anniegrossman.com/masterclass.
[Intro and music]
Annie:
So a few years ago near where I live in Manhattan, I saw a woman walking a dog with a shock collar, like a really big shock collar, bright, I think it was like bright yellow. And she had the remote very conspicuously in her hand. And maybe I shouldn't have done this, but I posted a photo of it, I think an Instagram stories of her with her dog and a shock collar. Like from the back, you couldn't really tell who she was or who the dog was.
Again, maybe I shouldn't have done this, but turns out she was a client of another dog trainer that has a studio, not that far from a school for the dogs in Manhattan. And I ended up taking the photo down but not before there was some back and forth in comments if I remember correctly on this photo. Maybe it wasn't in stories, maybe it was in the feed.
Anyway, there were comments, kind of along the lines of how we as quote unquote positive reinforcement trainers don't understand the importance of using all four quadrants of operant conditioning, and that — although I think this trainer referred to them as corners — that we, we really can't be good dog trainers unless we understand and use all four corners, AKA quadrants.
So operant conditioning, is the process of learning by consequence. If you do something, there is a consequence and the consequence can either be punishing or reinforcing. The consequence can be involving adding something or subtracting something. And if you're adding something, we call it positive. If you're subtracting something, we call it negative. If you're encouraging behavior it's reinforcement, and if you're discouraging a behavior, it's considered punishment.
So that's where we get positive reinforcement, negative reinforcement, positive punishment, negative punishment, which is often called the quadrants of operant conditioning. Although a few years ago, actually I read a really compelling article by Eileen Anderson of Eileen and Dogs, a blog I really love. And she made the case that it is misleading to refer to these as quadrants, that BF Skinner who codified so much of what we know and talk about when it comes to operating conditioning himself, never referred to quadrants. And that the reason it's misleading is because it leaves out extinction.
So you can get rid of a behavior using punishment, be it positive or negative, but you can also get rid of behaviors through extinction, which is basically when the consequence to a behavior is nothing. It's a deceptively complicated, slightly confusing part of animal training that it's taken me a long time to understand and think about, but it is important to know that extinction doesn't fit in neatly within the four corners as this other trainer said, or within the quadrants. So it can be misleading to talk about quadrants.
Anyway, if I'm talking about this exchange at all, it's only because I remember at the time thinking, okay, true that as a dog trainer I think it is important to understand how positive reinforcement, negative reinforcement, positive punishment, negative punishment, how all four of these things can be used to train a dog.
But ethically, I don't think we should be using positive punishment or negative reinforcement, positive punishment, positive in that you're adding something in order to discourage behavior, negative reinforcement being when we are encouraging a behavior because engaging in that behavior takes away something bad.
So positive punishment, dog training example. Let's talk about shock collars, for example, how you can use a shock collar in either of these ways. So dog barks, dog gets
shocked. A shock is the thing being added to the equation in order to discourage the behavior of barking.
Negative reinforcement. If you're using a shock collar that would involve say continuously shocking a dog with a remote collar until the dog drops something, or until the dog sits, or til the dog engages in whatever the behavior it is that you want. At which point, the dog stops getting shocked.
Now, I think what often doesn't get communicated is that there is a back and forth between the quadrant –If we're going to refer to them as quadrants and for simplicity, let's do so — between the quadrants that are like kitty corner from each other or, or diagonal from each other.
So if you think about waterboarding, for example, you are discouraging someone from being silent or from not telling the truth, by adding something unpleasant to the equation, like water, to the point where they feel like they can't breathe, that's positive punishment. And then if you are trying to encourage the behavior of them speaking, you're going to do so by taking away the bad thing, water that is causing them to feel like they're drowning, when they are supposedly telling the truth.
Now what's interesting is the same thing goes in the other direction with the other two quadrants, positive reinforcement and negative punishment. Now, I have never met a dog trainer who broadcasts the fact that they are a negative punishment dog trainer, which is totally understandable because negative sounds like a bad thing. And punishment sounds like a bad thing.
But guess what? I use negative punishment every day, and you can too, and it won't make you a bad person. In fact, it can often be a very mild way to discourage behaviors we don't want. A really great example of this is what we call the elevator game, which is something I have been doing with my new dog, Poppy, whenever I feed her her meals.
I have her food in. Usually I feed her in like a work to eat bowl. I really like the, actually the Northmate Green Feeder is my fave. So I prepare her food in that. She is sitting in front of me. I cue her to sit. I am standing up straight with the bowl in my hand. And as long as her butt remains on the ground, the elevator goes down. I bend over and bring her her food. The behavior of butt on the ground is being encouraged because it is equaling food coming her way.
However, the moment that her butt leaves the ground, the elevator goes up. I start to stand up. And this, my friends, is negative punishment in action. I am removing the food, taking the food away, that is why it is called negative, in order to discourage the behavior of her standing. That is why it is called punishment. It is discouraging a behavior.
Usually this results in her figuring out quite quickly, I think I should put my butt back on the ground, and the moment that she does — and by the way, I don't use a clicker or anything for this, I am just doing it with good timing. The moment her butt hits the ground, the elevator of my body starts to go back down to bring her her food.
And a tip: If you're just starting with a young dog doing this for the first time, I usually go through those last so-called floors, the last few inches from the air to the ground, pretty quickly if I'm working with a young dog. But as a dog gets better and better at this, you can get the food all the way to the ground and get to the point where you are standing all the way up again, and then cuing your dog to go get the food. But when you're just starting out, you want to make it pretty easy and you don't want to push them quite that hard.
Now you could also do this, for example, with a dog who is trying to burst through a door or through a gate. As long as the dog has four feet on the floor, the door opens, or the gate. If the dog starts to jump, the door then closes. Or of course you could do this with a sit too. As long as the dog is sitting, the door opens. If the dog stands up, the door goes in the other direction.
An example of a negative punishment with people would be fines. When, when you take money away from someone in order to discourage a behavior or grounding a kid, for example, similar idea when I have to pay sales tax to the, to New York City and New York state every quarter, which is like my least favorite activity, part of what motivates me to do it on time is I get a little bit of a credit. Like they give you like $200 or something if you are on time, positively reinforcing the behavior of collecting this money and giving it to the government. And if you are at all late, they fine you. So again, you can see that diagonal between positive reinforcement and negative punishment.
I was trying to think of other examples of negative punishment in human life. so I pulled out one of my favorite books, I guess it's actually like a textbook. I think it's called Behavior Principles in Everyday Life by John and Janice Baldwin. And there's actually only like two paragraphs on negative punishment in this three or 400 page book. But they sure are interesting paragraph. So I thought I would actually just read them:
Punishment by subtracting. The second type of punishment occurs when an operant leads to the loss or subtraction of a positive reinforcer. When people drop or lose things of value, the loss produces a type of punishment, often effectively suppresses future clumsy behavior. Have you ever lost your wallet full of money and identification cards?
You know what this is, this is me. I'm not reading anymore. You know, what I like about this first example is that it points out that, we're learning because of our environment and our own actions all the time. So much of dog training comes across as like what we do to dogs. But this is just a reminder that punishment, reinforcement, not necessarily anything that we do to dogs or that is done to us. It just occurs in life. Okay. Where was I?
Have you ever lost your wallet full of money and identification cards? The loss subtracts from your happiness and punishes the negligent behavior that led you to misplace it in the first place, making you less likely to misplace it in the future. We stopped putting quarters in the candy machine after losing several quarters in a row and getting no candy.
Clumsy, inconsiderate, or crude behavior in social interaction may result in a loss of friends or possible new acquaintances. When a person says something insensitive at a party and several of the listeners tuned turn to join other conversations, the loss of social reinforcers punishes the behavior of saying insensitive things in public.
The loss of positive reinforcers subtracts from happiness. The loss of a driver's license punishes, fast driving, and often suffices to suppress speeding in the future. The loss of love when a good relationship goes on the rocks punishes and may well suppress some of the inconsiderate behavior that damage the relationship. Afterward, a person may think over and over, ‘I'll never be inconsiderate again. A beautiful love is so rare that I never want to lose love again.’
Parents can use punishment by subtraction as an effective means of behavior modification for helping children overcome problem behavior. For example, thumb sucking is a frequent and persistent response that can cause problems for many children. Thumb suckers are often teased and treated as if somewhat immature. Years of thumb sucking can also cause serious misalignment of the teeth.
A study at the university of Vermont showed that thumb sucking could be totally eliminated by the second type of punishment, even in children as old as six to eight years. All three children in the study had been persistent thumb suckers since infancy. The parents began an experimental program for 10 to 20 minutes each day at bedtime. At first, the parents read favorite stories to their child, no matter whether or not the child did thumb sucking. Figure three to seven shows the data from six year old Heidi. And it is clear that story reading did not eliminate the undesired behavior.
Then a period of punishment by subtraction was initiated. The parents stopped reading when the thumb touched or entered the mouth and only resumed reading when the thumb sucking ceased. There was a rapid suppression of thumb sucking as can be seen in the data at period B. When the punishment was temporarily discontinued and the parent read stories, whether or not the child sucked on the thumb, the thumb sucking reappeared. When the parents resumed punishment by subtracting their reading when they noticed them sucking the response again declined. Thumb sucking was eliminated in all three children.
Isn't that interesting. I find this book endlessly endlessly interesting. You can get a used copy of it on Amazon for just $4. How incredible is that? I think I'm gonna leave it there. I might call this episode in defense of negative punishment. Although I probably will not start calling myself a positive reinforcement and negative punishment dog trainer, I suppose I am.
