03 AprEpisode 47 | A busy person’s guide to operant conditioning
Positive Reinforcement, Negative Reinforcement, Positive Punishment and Negative Reinforcement: You may have heard these terms thrown around, but they're often misused or little understand. Together, they describe four different ways all animals learn by consequences-- aka, Operant Conditioning. In a rush to put up a podcast episode while her young baby naps, Annie challenges herself to describe each "quadrant" of Operant Conditioning in two and a half minutes.
Subscribe: iTunes
Transcript:
Annie:
Hey everyone. Annie here. And I'm with my three-month old cohost who is being quiet right now, but she may have something to say shortly. And because I am a full-time mom in addition to running School for the Dogs and in addition to doing this podcast, I am trying to be more efficient and how I do everything. Because although I have seven hands and four brains, it's hard to get it all done. So for that reason, I am challenging myself today to keep this episode short. And the thing I want to talk about today is operant conditioning, which really has four major parts. So I decided to see if I could describe each of those parts in two and a half minutes. Making all of operant conditioning, something that can be contained in 10 minutes. And let's see. Let's see if I can do it.
But first, what is operant conditioning and why do I want to talk about it? Well, there are basically two ways that all animals learn. There's two kinds of conditioning. Conditioning is a term that's synonymous with learning. And those two ways of learning are operant conditioning and classical conditioning. And if you've listened to this podcast before you've heard me talk about these two kinds of learning, I think that they're very important to understand when you're trying to think about how your dog is learning, what you're trying to train him or learning what you're not trying to train him. And I am well aware that most dog trainers don't talk about the science of dog training as much as they talk about specific problems and how to solve those problems. And while the trainer may or may not understand these two kinds of conditioning is something that you as the person who's hiring the dog trainer may never know.
And I was thinking about this. Last weekend, I went to Clicker Expo with a bunch of people on my staff. Clicker Expo is held three times a year twice in the States, once in Europe and it's like the major gathering of people who are geeky about dog training. It's run by the same people who run the Karen Pryor Academy, which is the dog training school I graduated from. Some of the best trainers in the world speak at Clicker Expo. There are labs, there's a trade show and it's just a wonderful gathering place of positive reinforcement trainers where they do talk a lot about things like operant conditioning.
Anyway, at the end of one panel I went to, I heard a woman talking to someone and she said, “you know, all of this conversation about the science is so interesting and so fascinating, but my boss would kill me, would kill me if I ever tried to explain this kind of science to my clients. Their eyes would just roll into the back of their heads and they wouldn't pay attention to anything that I'm saying. So I guess the real trick is trying to figure out how to communicate this stuff to people without letting them know that there is science involved.”
Now I totally get that because the word “science” is scary to a lot of people. And if science or topics relating to science aren't communicated in the right way and aren't made to be applicable to your life, I definitely know that, you know, your brain can turn off. I had science class from what, like fifth grade through 12th grade and I learned pretty much nothing, which is scary and embarrassing but true. I learned, I mean there's certain words I remember Xylem and phloem have something to do with celery. The periodic table is a thing. Newton was a guy. I'm exaggerating but really only slightly. It's amazing how little I took away considering all the hours that I spent in science classes in a very good grade school. Honestly, the main thing I remember about all of the science classes I ever took was once in like fifth or sixth grade, there was this girl named Nina who sat behind me and did origami the whole class. For some reason she got, like, a special dispensation and she was allowed to do origami. And I was, I always liked to draw and I would like to doodle during class and I was always getting into trouble for doodling, but she was somehow allowed to do origami. And one day in middle school science class, I just lost it. And I turned around and started yelling at her ‘cause I just like couldn't deal with the injustice.
And that's my main memory of all those years of science, anyway. Now, of course, I think science is the coolest thing. I don't think I'm any good at science, but, whereas, I used to think, “I'm not good at science and therefore I'm just going to ignore science.” Now I think you know what, even the best scientists in the world are still figuring things out and probably feel like they're not that good at science because science is not one thing. It's kind of everything. And the science of behavior, to me, is the most relatable science because we are all animals who are behaving all the time. And our ability to learn relates to our ability to adapt to different environments and which certainly has to do with learning and is why we've been such a successful species and why dogs have been such successful species. And it's a big reason why dogs and humans are two of the only species that have ever existed that aren't yet extinct. I did an early episode called Our Canine Cousins, which I will link to in the show notes, about all the ways that humans and dogs are exactly alike. And that is one of the ways in which we are like 99% of all species who have existed since there has been life on earth, have gone extinct. Humans and dogs are in this tiny club of species that are still around. And I would certainly argue that our ability to learn really well thanks to operant conditioning is a major reason why we are still here.
So anyway, yes, science is hard. Science is complicated. But I think that this little bit of science called operant conditioning is surprisingly understandable because it's something that we are all experiencing pretty much all the time.
And you know, my mom all growing up, she used to say, “I don't really care what you're learning. I just want you to, I just want at school, I want them to be teaching you to learn how to learn.” And recently I actually said to her, “you know, you were always talking about wanting me to learn how to learn. But actually that's the one thing that kids don't have to learn.” We think about learning as something that happens in school, in a classroom when you're taking a test or reading a book, but really learning is happening every moment, every day, and from the time that you're born, you're learning. So I said to her, you know, “the one thing that they didn't have to teach me at school was to learn how to learn.” Actually, maybe the one thing that they did teach me at school was to not want to learn to want to go home and veg out in front of Saved by the Bell and play Nintendo. But you know what those, those things involved learning as well.
Anyway, I really believe if you can get your head around operant conditioning, which I'm about to explain it can really help you think about why your dog is doing, what your dog is doing and it can help you think about why you do what you do and why you don't do the things you don't do.
Okay. Just a little quick background on operant conditioning. Operant conditioning has always existed, but it was codified by BF Skinner in the mid 1900. I've talked about BF Skinner many times before on this podcast. I actually interviewed his daughter last summer and he was a psychology professor at Harvard among other places and he might be best known for a lot of the work that he did with pigeons and rats. But he was also a dog trainer. He did some fun training with his dog, which I mentioned in the podcast with his daughter. And at the end of the day though, his main concern was training humans. And when you talk about training humans, it sort of has a nefarious ring to it. But he wasn't evil at all. He really felt that we could lasso what we know about human behavior and how all animals learn in order to make people happier, make the earth a better place. He mostly wrote nonfiction, but he wrote a really interesting novel called Walden Two, which is about a utopia that is run basically by a behaviorist. And when it came out, it was referred to as fascism without tears. And there were also our reviewers who said it was kind of as if his idea was to treat all people as if they were dogs in a training class. And although I think that that was, sort of, a criticism of the book, I actually think it's kind of accurate. It's like what would the world look like if it was run by positive reinforcement dog trainers or applied behavior analysis?
Okay. So our conditioning is really just a fancy term for learning by consequence. And it's often taught by talking about it in terms of four quadrants, which if you're a visual learner, I think can be helpful. I think about it as a sort of picture like an X, Y axis with the positive stuff being above the X axis and the negative stuff being below the Y axis. Now when we talk about positive reinforcement, which is part of operant conditioning, the positive, as I've mentioned before on this podcast, does not mean happiness, smiley faces, heart, stars and flowers. It's positive as in mathematically positive so that's why we're putting it on that part of the axis. Negative is going to mean taking something away which is why it's on the bottom part in the axis. Now reinforcement is going to be on the right side of the Y axis and punishment is going to be on the left side of the Y axis.
Reinforcement, and when we're talking about operant conditioning, simply means encouragement. If behavior is reinforced, it means it's more likely to happen again and punishment is just the opposite. It means discouragement. If the behavior is punished, it means it's less likely to happen again. Now simply stated, every time you behave in any way, there is some kind of consequence and that consequence falls into one of these four quadrants. Now just to be clear, I should say that there is a fifth element to all of this which is called extinction, which is, kind of, when nothing happens. So, kind of, like every time you do something, your behavior is either reinforced, in which case it's more likely to happen again in the future or it's punished, in which case it's less likely to happen again in the future or it's extinguished, which basically means your behavior has no effect, nothing happens, there is no consequence. And there are behavior nerds who disagree with describing operant conditioning in terms of quadrants because there's four quadrants, there really is no place there to explain extinction, which is very important.
But it's also kind of complicated. I think, I don't think I really appreciated or fully understood extinction until I started training chickens, which I did four or five years ago. Up until that point, I think I got the basic idea, but I didn't appreciate how powerful and interesting extinction is. But because I am trying to keep this podcast episode short and not overcomplicate things, I'm going to save the topic of extinction for another episode. And instead just focus on the four, let's call them quadrants of operant conditioning: positive reinforcement, negative reinforcement, positive punishment and negative punishment.
So we're going to start with positive reinforcement. I'm setting my timer here for two and a half minutes.
Okay. So positive means we're adding something to the equation. Reinforcement means that we're encouraging the likelihood that a behavior's going to happen again and positive reinforcement is affecting you all the time. In fact, if you think about everything that you did today, if it was something that you have ever done before, that means that most likely it was reinforced in some way. And if it was positively reinforced, that means you did it because it yielded something good, something good happened. So if you go to work and then you get paid, the money is a thing that's added to the equation and it encourages the likelihood that you are going to keep going to work. Dogs are positively reinforced by food, of course. We use a lot of food and training, but really anything that your dog likes could be used to reinforce a behavior, attention, the opportunity to go for a walk, to chase a ball.
And an important thing to remember about positive reinforcement, I think, is that a lot of the times when we're trying to discourage a behavior, we end up positively reinforcing a dog with attention, right? So your dog jumps up on you. You say, “no, don't do that.” Well, you know, your attention might have been positively reinforcing to the dog and you can tell whether or not it was, if the behavior keeps on happening. If the behavior doesn't happen anymore, then you have effectively punished it, but we haven't got to punishment yet. And I think that that tends to happen a lot. So if there's something that your dog is doing that keeps happening that you don't like, you need to figure out how is that behavior being reinforced. And if it's being positively reinforced, that means that something is being added that is encouraging the behavior.
It's also important to think about the fact that, you know, what might positively reinforce a behavior for one person might not be positively reinforcing to someone else, which is why it's important to know what's rewarding to your subject. But basically, you know, just putting one foot in front of the other. If you put one foot in front of the other and it makes you go forward, then that behavior has been positively reinforced and you're much more likely to take another step. It can be that simple and that minor.
Oh ok, got to two and a half minutes.
Next up is negative reinforcement. So negative, you're taking something away. Reinforcement- in order to encourage the likelihood that a behavior is going to happen again. And this can be a tricky one to think about, but I think a good example in human life is the beeping noise that you hear in your car when you don't have your seatbelt on. You put your seatbelt on and the beeping noise goes away. So basically the behavior of putting on a seatbelt is encouraged because it makes the annoying thing, the bad thing go away. You wake up in the morning because your alarm goes off and you smack it to make it stop. But you have gotten up because it has made the annoying thing, the sound of the alarm go away. You have a headache, you take a pill, makes the headache go away. The behavior of taking the pill is encouraged because it makes the bad headache go away.
I had a client once who said, “well, can you give me an example of how you would use negative reinforcement with a dog?” And I said, “no, I can't because I would need to be putting my dog in a situation where he'd be experiencing something bad or it would be experiencing some kind of pain and then I'd need to make it stop. And that's not something I would want to do to a dog.” But there are trainers who train using negative reinforcement by, for example, pinching a dog's ear until the dog drops something, that's using negative reinforcement or shocking a dog via shock collar until the dog sits. That's another way that people often use negative reinforcement.
In my dog's life, there was one time where I saw negative reinforcement effecting him, although I wasn't using it on him on purpose. He got stuck under my uncle's porch and he was shrieking. He was really scared and uncomfortable. He managed to get out through a small hole on the side of the porch. And I would say that's an instance of negative reinforcement that, in the future, were he ever again to get stuck under the porch, the behavior of exiting in that one small spot would probably occur again because it made the bad thing of being under the porch go away.
All right, now we, now we are going to go to the other side of the axis and talk about punishment or discouraging behavior. Right? Let me step my timer again. Okay.
Let's start with positive punishment. Positive punishment is a very strangely termed thing because, you know, positive sounds like it's something good, but again, we're thinking about positive in terms of math here. So we're adding something to the equation in order to punish it in order to discourage likelihood that a behavior is going to happen again. So if I punched you, I am adding my fist to the equation of your face in order to discourage you from doing whatever it is you just did. Although again, like I talked about when I was talking about positive reinforcement, you have to think about what is punishing, what is punishing to your subject because some people might like being punched in the face. But basically whenever you are doing something that is causing discomfort, causing pain, causing anything disliked by your dog or whatever animal you're working with in order to keep them from doing something, you are effectively using punishment. But, like I mentioned with positive reinforcement, I think there's a big mistake that people make a lot of the time where they think they're punishing an animal when they are actually positively reinforcing it. They think they're punishing the animal by yelling or by whatever. But you know, if you actually look at whether or not the behavior is continuing, you can tell whether or not it has been reinforced or punished. If it has been punished effectively, the behavior should happen less and less and less to the point of it not happening at all.
There are a couple of really big problems with using punishment. One is that animals tend to become desensitized to punishment. So you have to make your punishment greater and greater. I know I experienced this as a little kid. I used to get spanked and, at some point, it just didn't really care about being spanked anymore. And I would say, “what are you going to do spank me?” You know, that's, you know, what are my parents going to do, start beating me?
You know, when we're trying to use punishment, we often lead dogs to make the wrong association. So, you know, you yell at your dog when you come in and see that your shoes or your shoes are chewed up, but your dog now thinks he has been punished for
Timer goes off –done almost done with what I was going to say.–
Your dog thinks he's been punished for running up to you when that was not what you were going for. So those are just some of the problems I've encountered with seeing people attempt to use punishment with their dog. You think you're punishing your dog for one thing, but your dog might have a totally different idea about what you're punishing your dog for. Another thing, and I know I'm over my time now, but these are my rules, I'm going to break them, God Damn it. I know in my life I can remember punishments that I've received a lot more than I can remember what I was being punished for. And again, that comes down to your dog making what might be a bad association because he's making the association of the punishment with the punisher. And that memory might loom much larger in your dog's mind than remembering whatever the behavior was that he was punished for that he shouldn't be doing anymore.
Okay. Last but not least, we have negative punishment. Let me reset my timer here. I'm going to give myself two minutes on this one instead of two and a half cause I went a little bit over in the last one.
All right. Negative punishment. We are taking something away in order to discourage a behavior. Human example of negative punishment would be grounding a kid. You're taking away their freedom, their car keys, whatever, in order to discourage the behavior of, you know, staying out too late or whatever it was.
What's interesting about negative punishment is there are these crisscrosses that happen. So if you're using positive reinforcement, you are probably gonna end up using some negative punishment. Even if you're not meaning to and same thing with the other crisscross of positive punishment and negative reinforcement. What do I mean by that? Well, let's say you're playing what we call the elevator game. That's an example of a place where you are using negative punishment. The elevator game is when you're, you start to put a food bowl down while your dog's butt is on the ground and if your dog's butt starts to come up before the food bowl is all the way on the ground, you start standing up again. You're taking away the food in order to discourage the behavior of standing up. But then you're bringing the food back down when your dog sits again, so at that point, you're using a positive reinforcement. So if there's sort of like a back and forth that's always happening in life.
Another example would be, you know, your dog jumps up on you and you turn your back on the dog. You're taking yourself away from the dog in order to discourage the jumping, but then you turn around again in order to reinforce the dog when your dog is not sitting, then you're using positive reinforcement. And in the other direction, you know that alarm clock example I gave, you know you are getting up because it makes the sound of the alarm stop. So that is negative reinforcement-the behavior of you getting up is encouraged. But the behavior of your sleeping is discouraged, is punished because of the thing that's added- the annoying sound of the alarm clock. So you can see there's always like a seesaw arrangement that's happening when you're thinking about reinforcement and punishment. And the direction we're always trying to go towards is using positive reinforcement and negative punishment rather than negative reinforcement and positive punishment.
All right. I did it. Operant conditioning explained in about 10 minutes.
And in lieu of a Woof Shout Out or a Fun Dog Fact today I wanted to just take a couple minutes to respond to some comments I got on a couple of recent episodes, although I think that, actually I know that the people who sent me these comments didn't actually listen to the podcasts that they were commenting on, they were just responding to the descriptions of the podcast so I doubt that they will listen to my response here. But nevertheless, I thought I would offer up some thoughts on these things.
The first comment I wanted to respond to was from a cousin of mine who I haven't heard from in years. I didn't hear from this person when I got married, when my father died, when I had a baby, nothing, nothing, nothing. And then I won't say if it's a he or she, but this person saw something about the episode that I did about Marie Kondo and I got a couple of really negative responses from people who said that they had no interest in listening to my episode because they were so disgusted by the description of the episode. Fair enough. This is just one that I thought I would read, but most of them were saying, well actually not true- I got two kinds of responses. One kind of response was saying basically, how dare I say anything bad about Cesar Milan who is mentioned in the episode and also that I was taking a Western view of Marie Kondo, who’s practices were, are rooted in ancient Japanese tradition which I am taking a Western point of view because I am from the West and I, if as anyone who listens to the episode knows, I actually think Marie Kondo is great in so many ways. The point of the episode was talking about a lot of the superstitious things that she does. And you know, at Clicker Expo I heard somebody talk about the episode of Friends where Phoebe thinks that she's controlling the TV, the TV power by blinking her eyes when it's actually Monica on the other side of the room playing with the light switch. That's a good example of superstitious behaviors that people think are linked to a behavior that really has nothing to do with anything.
Anyway, my criticism of Marie Kondo was about her suggesting that you talk to your books and your clothes, etc. And the criticism that I got from this cousin touches on that too. This cousin wrote
“just wonder if making the religious renounce their beliefs because empirically Moses could not part the Red Sea is necessarily important. If your goal is to help people train their dogs and enable both to have enriched experiences that they are willing to pay you for then I'm not sure you have to dispel all the woo. Energy, as Cesar puts it, is an adequate and not wholly inaccurate way to describe the communication between human and dog. This is what animals us included. Tune into. And yes, energy can be described as a bunch of physical and chemical mechanisms, but the simplification still resonates, right?”
So anyway, I wrote this cousin back and I said, “thanks a lot for telling me how I should be a dog trainer who's going to get people to pay her when you haven't reached out to me for any other reason in my life in a decade.”
But you know, as far as whether it's Cesar Milan talking about energy or Marie Kondo, talking about speaking to your clothing or Moses parting the Red Sea. First of all, let's, well, let's talk about Marie Kondo and Moses first. I don't think it's bad to do these things. There's certainly nothing wrong with believing that Moses parted the Red Sea or that you need to talk to your socks. Or any, anything, you know, he mentions religion. Anything that makes you feel good, whether it's eating the body of Christ or you know, snipping off the tip of your baby's penis or, you know, lots of things people do in the name of religion, which could be called superstitious. I'm not really judging as much as I'm saying that these are not the most effective ways to change behavior and, you know, talking about energy, sort of like the same thing.
If you think about the principle of parsimony or Occam’s Razor, you know, we're trying to find the simplest answer to questions. We're trying to, in this case, talk about changing behavior with, by taking the fewest assumptions, sort of, taking the most direct path. And I believe that an understanding of the science of behavior and understanding operant conditioning and understanding classical conditioning and the law laws of learning are just a clearer path that involve fewer assumptions then attributing behavior change to energy, which could be defined differently from one person to another. And another thing when you think about learning, so much learning has nothing to do with anyone. I mean, of course, you learn from a book for example, and you learn from an online class. Are you learning because of the energy that was infused in that book or the energy of the person teaching the class? I mean, I don't think so, but even beyond that, all animals learn how to live in their environment or I mean the animals that have survived and flourished. All animals have learned how to adapt to their environment and how to do what works and how to not do what doesn't work. And all of that doesn't have to do with the energy of a teacher. The teacher is the environment.
Anyway, that's my short response to those criticisms which I received. And honestly what I regret is that my descriptions of the episodes were so off putting to these people that they didn't listen to the episodes at all. That certainly doesn't do the job of a description which should be enticing. So the behavior of writing descriptions in that way I guess has now been punished and I will be less likely to do that again in the future.
The other episode that I got some comments about was the episode I did recently where I interviewed Cherrie Mahon, owner of River Valley Doodles. I had some people who wrote me in disgust saying that they couldn't believe I was doing anything to promote breeding. And this is something I could talk a lot more about, but I just wanted to say something brief about how we get dogs into our lives. I do think that going to a shelter is the best way to get a dog right now. And those who listen to the episode will note that Cherrie said the very same thing, but that said, we should aspire to a time where there are no dogs in shelters which would require, you know, a big spay/neuter effort, which is happening. But dogs are still going to mate.And if dogs are gonna mate and people are still gonna want to have dogs in their lives, I think that it's worth thinking about how we can produce dogs who are healthy and are raised in the right way and don't have genetic predispositions that are going to cause problems and break the hearts of the people who love them later on. And yeah. And they should be going into homes that are vetted by people who care about the long-term wellbeing of the dog.
So ideally, yes, I think we should get all dogs out of shelters, but I think that in tandem we should also be thinking about, if we ever are going to achieve that goal, it's good that there are people now who are working on figuring out how to breed dogs that are healthy, that have genetics that are, that are good, that are being tested, that are not too inbred. And Cheri is someone who is an example of a breeder who's really putting in the work to create dogs that are going to be great pets and that are going into good homes. So while I am all for rescue and going to shelters and working with rescue groups, if you're looking for a dog, I also am a big supporter of people like Cherrie who are breeding dogs, specifically breeding mixed- breed dogs. She breeds Doodle mixes and who are doing it in all the right ways.
If you would like to positively reinforce the behavior of me doing this podcast, I would love it if you would leave a review on iTunes, tell your friends, post in Instagram stories or in your feed or on Facebook that you're enjoying this so that more people can learn about it. That would all be great, would certainly be rewarding to me. And as always, I really appreciate everyone who reaches out via email or on DM and Instagram. If you have any questions about dog training you can email podcast@schoolforthedogs.com or you can call and leave a message (917) 414-2625. I will definitely try to answer your question and put it.
Links:
