12/27/2024
Copied from Chad McKin:
This is the kind of thing I'm sharing on other platforms that actually send my work to all the people who subscribe to my stuff.
**** ABC is Wrong (Kind Of) ****
Not long ago, I saw a video shared by a shelter. It featured a blue pittie who leaped up and grabbed the leash whenever anyone went to walk him. The shelter described this as dangerous behavior and wanted to find a way to stop it. The narrator said they'd tried almost everything, including medication, to solve the problem. When nothing else worked, they decided to try punishment.
They chose to use a shaker can.
The narration made it clear that the shelter wasn't happy about resorting to the shaker can. I agreed it wasn't an ideal solution, but not for the same reasons.
I have no interest in shaming or even judging the rescue in question. Just the opposite in fact. They have an institutional aversion to using any form of punishment, and they bent that rule to help a dog when they thought it was their only option. That's way more morally responsible than declaring the dog unsalvageable and giving up. You can't say you've "tried everything" until you've actually tried everything. I promise they got plenty of backlash over this, too. The decision to not only use punishment but to make a post about it on social media was a courageous one.
They should be applauded for that.
I am not personally opposed to the thoughtful use of punishment at all. This application wasn't unfair or unkind, and it was somewhat effective. I just believe it was premature.
Watching a video and playing armchair quarterback is always tricky, so I might be wrong. But I feel as if I've met 100 versions of that dog, and the dog just needs some good play.
Dogs are predators, and deep in their hearts, they want to hunt, chase, and kill. Because of selective breeding and lots of management, most dogs are fine without actively expressing that drive. For certain dogs, though, the drive to hunt or do things similar to hunting is overpowering. If there isn't a way to express that drive, these dogs are likely to engage in lots of stress-related behaviors. This particular dog was super clear about it.
He kept grabbing the leash.
He wanted to play tug!
A good game of tug, with the proper rules in place for a few minutes a day, will quickly make this behavior manageable.
No shaker cans necessary.
Here's what confused me about this.
Several years ago, I committed to working with every dog in the softest way I could and still being effective.
I figured if I was serious about using as little compulsion as I could get away with, I'd have to really look into what people who never used it were doing and take as much from them as I could. So, I learned and practiced many of their strategies until I became competent with them.
I don't know if there are people out there who can solve every behavior problem without punishment, but I do know that I can't. On the other hand, I also know that I can solve a lot more of them that way than I ever suspected.
So, when I see a force-free organization missing something like providing biological fulfillment, it confuses me.
If a trainer who doesn't limit himself to force-free methods can immediately see a reward-based solution to the problem, why didn't they see it, too? What am I missing?
They obviously care about the dog a great deal and are putting in a lot of extra effort to give him the best chance at a great home. They clearly want to do things in the kindest way possible. But it seems like they missed a very easy and effective solution that the dog will absolutely love! Why is this so obvious to me but not even on their radar?
Well, I have an idea.
This isn't the fault of the shelter workers or even the shelter management. I think the issue is systemic and baked into the way trainers view the world. The history of behavioral research may be at the heart of it.
These days, there are lots of scientific disciplines that study behavior. But in the early days of dog training, only two counted when it came to animals. They were Ethology and Behaviorism. Ethology is the study of animals in their natural environment. Behaviorism is the study of changing behavior and is usually practiced in a controlled environment.
Ethology depends a lot more on observation of day-to-day activities than experimentation. It has given us the ideas of Social Hierarchies, Dominance, Imprinting, Appeasement Behaviors, etc. (social hierarchies and dominance are real but very misunderstood, especially as they relate to dogs)
Behaviorism depends more on experimentation in changing behaviors under specific conditions. It has given us the ideas of Operant and classical conditioning, Counter Conditioning, Desensitization, etc. Really, most of the jargon dog trainers use is founded in Behaviorism.
It's important to understand that these two disciplines not only had very different ways of describing and analyzing behavior, but they also began with different assumptions.
The ways I learned to train dogs had much more to do with Ethology than Behaviorism. I believe this is largely true for most of us who came up more than 30 years ago. Behaviorism didn't become mainstream in the dog world until a little later, with the rise of "science-based" training.
When I say my early learning was more based on Ethology than Behaviorism, I don't mean my books were quoting well-known ethologists and referring to researchers' field notes. I mean, they were generally approached from an observational point of view. It was easy to observe how dogs set boundaries amongst themselves. It was also easy to observe how dogs responded to events in their environment and interacted with different people in different circumstances. We can see how dogs spend their free time when not directly interacting with people. It was obvious to most people that dogs were social beings and were largely driven by social desires. Whether Konrad Lorenz's name was ever invoked or not, dog trainers and dog owners were basing many of their ideas on what could be observed through everyday interactions with domestic dogs.
The point is we assumed dogs had their own goals and ideas and that those goals and ideas were intrinsic to their choices. So you would get statements like this passage from the first dog training book I ever owned, written in the 1960's:
"It is claimed that some animals are more intelligent than the dog but, by any standard of comparison, in emotional stature the dog is the closest creature to a good person. This means that a dog needs and deserves a pleasantly rounded life. True working dogs enjoy opportunities for this expression in the jobs for which they were found suited, but even the fullness of that gratification doesn't lessen an even deeper need. Above all other animals, the dog has a need to receive as well as give honest affection."
Some of the assumptions made here might not be provable, and some may be dead wrong. I don't know that there's been a ton of research on whether dogs "need" to give "honest affection." That's beside the point. I didn't pull that quote to talk about the truth of it. I'm pointing this out because it talks about dogs as emotional creatures motivated by those emotions. The author doesn't bother taking time to prove the case either. There is no need to prove what is obvious.
Behaviorism took a very different approach to looking at behavior. As the name implies, Behaviorism concerns itself specifically with behavior and how to change it. In the early days, there was no way to look inside an animal to see the decision-making process, so they limited their research to observable behavior. The process that occurs between "stimulus and response" was considered unknowable. Researchers weren't interested in speculating about what was going on in that black box. Instead, they focused very intentionally on what could be observed. In this environment, the goal was to cause, change, or eliminate specific behaviors as isolated events. Eventually, Behaviorists would declare that "thoughts don't affect behavior." B.F. Skinner, who might be the most influential Behaviorist to date, didn't seem to believe much in genetic predispositions either. He believed that through proper reinforcement and punishment, he could take any infant and eventually create a great doctor, scientist, or anything else.
In the nature vs nurture debate, he was on team nurture.
Behaviorists weren't really interested in species (or breed) specific behaviors. They were more interested in generalized, universal principles for changing behavior. They did not consider things like instinctive drives or biological fulfillment. The reasoning was that all of that stuff was speculative at best. Why guess about motives when we can change behavior predictably without guessing? We can measure whether a behavior becomes more likely or less likely. We can experiment specifically on what "consequences" most effectively create behavior changes.
Scientific methodology could be more rigorously applied because it was practiced in a laboratory setting. What we couldn't do was observe motive. So, the behaviorists didn't bother with guessing about things like that. They did just fine changing and creating behaviors without all that nonsense.
Using such a narrow focus proved beneficial in many ways. Many of the strategies I use every day are the result of the hard work of behaviorists over the years.
Because of the scientific rigor of laboratory experiments, modern trainers tend to base most of their theories on the studies done by Behaviorists such as B.F. Skinner, Marian, and Keller Breland were huge pioneers in the field. The work of ethologists is less influential these days.
Given the influence of Behaviorism on dog trainers, it's no wonder that most trainers tend to see behaviors as isolated events that can (and should) be adjusted through rewards or punishments.
One of my mentors, Dick Russell, used to say "Everything in the world we know about dog training can be summed up in two sentences. If just as or just after a dog does something, something good happens, the dog will do that again. If just as or just after a dog does something, something bad happens, the dog won't do that again." Simple right? And it's true as far as it goes. But it's missing something.
It doesn't address the obvious question: Why did the dog do that thing in the first place?
In The Culture Clash, Jean Donaldson describes a scenario where the dog wants something out of the fridge and tries to get it by staring at the refrigerator, then scratching at the door, and eventually giving up because neither works. She asserts the dog is not "working this out logically in his head" but is randomly trying things to see if they work. In the mind of the Behaviorist, there is no intelligence behind the behavior; there is only an intrinsic reward and the desire to obtain it. This idea has been more or less debunked, and most dog owners have always known this is not true. But if you accept the notion that thoughts don't influence behavior, you can't really believe in thoughtful choices.
The mantra of old school Behaviorists is "Consequences drive behavior."
In that framework, it makes perfect sense that most dog trainers focus primarily on the specific behavior they are trying to create, reduce, or eliminate. If you talk with trainers for some time, you'll likely run into the ABC idea.
ABC stands for Antecedent, Behavior, and Consequence, which is the fundamental way behaviorists view behavior. Even if they never describe it this way or have never heard of the idea, ABC still accurately describes the way most trainers approach behavior.
In Neuroscience related to habits, they use Cue, Routine, and Reward, which more or less follow the same structure with a slightly different focus.
Antecedent (cue) refers to the circumstances that incite a behavior (routine), which produces a consequence (reward). It's a neat little package for troubleshooting or changing behavior.
For example, the presence of another dog would be an antecedent to lunging (behavior), and the other dog's owner taking the dog across the street would be the Consequence.
If that Consequence pleases the dog, that behavior is likely to be repeated. So, trainers will likely focus on changing the behavior through consequences. For example, they might give the dog a leash correction for lunging. Or they may try to interrupt the sequence and cue a "sit" or a "watch me" before the dog reacts, offering a reward for doing something other than lunging. A trainer might try many things, but to this day, most of them will be built on this structural framework.
In the video I was talking about earlier, this was the thinking that got them to the shaker can: antecedent (someone shows up with a leash), Behavior (grabbing the leash), and Consequence (loud noise which discourages leash grabbing).
And it did seem effective. However, the dog still looked frustrated. He made a better choice, but he was still thinking about the wrong one. His goal hadn't changed, only his behavior. After a few days or weeks of this strategy, I suspect he will return to his previous behaviors. I could be wrong, but it wouldn't be surprising if it did. The dog just had too much internal conflict.
Under the ABC model of behavior, there is some room to consider things like overall health and biological fulfillment, but they aren't typically emphasized. Skinner's work was primarily focused on the "consequence" part of the equation. And when it comes to things like emotional health and mental resources, those ideas would require too much speculation.
Even though many of the tenets of behaviorism remain valuable and helpful to this day, many of its foundational premises have been changed by later scientific developments. Neuroscience has been able to pierce that black box to a large extent, and we now know dogs (and all mammals) have very similar emotions to people; they have goals and motivations beyond reinforcement histories, and they are intelligent and complex. In Neuroscience, the mantra is not "Consequences drive behavior" but "Emotions drive behavior." Which makes a lot of sense. Consequences will only affect behavior if they affect emotions. Otherwise, they are "inconsequential." If a piece of kibble doesn't excite a dog, he's not going to change his behavior to get it. With no emotional change, there is no reason to change a behavior.
We understand so much more about emotions and how they affect behavior than Skinner could have imagined. There's still a lot left to learn, but we aren't operating in total darkness. Understanding that Emotions Drive Behavior gives us a fresh starting point for looking at behavior and a broader way to approach the ABC model. We can look at things like emotional responses and mental fatigue as parts of the antecedent cocktail.
How the dog responds to the antecedent is not merely the result of past experience, nor is it simply a random response. The dog's response depends on a complex web of influences, including instinct, drive, mood, health, biological resources, and core affect. The antecedent doesn't illicit behaviors per se. It changes how the dog feels. It's those feelings that start the behavior change.
One of the big mistakes people make when thinking about these things is focusing on the moment instead of affecting changes down the line. Pavlov didn't condition the dog with one bell. It took many reps for the dog to make the association. We aren't merely trying to manage the dog in the situation but changing how he feels about similar situations in the future.
Something Skinner got really wrong was the idea that learning is an event, not a process. That was a foundational concept of behaviorism, and it was baked into all the literature and experiments on which Behaviorism is built. It's an idea that has messed up a lot of trainers. If they don't see a change quickly, they conclude their approach isn't working.
But that's often not true. Once we begin to see learning as a process, we are confronted with the idea that things can change quite a bit internally before we see a behavior change.
I heard a great analogy once. Imagine you're in a room, and the room temperature is 30 degrees Fahrenheit. In the center of the room is a table with a single ice cube. On the wall is a crank you can turn to raise the room's temperature. If you crank that crank really fast and raise the temperature to 31 degrees, the ice cube will remain unchanged. You might conclude that your efforts are having no effect, and the ice is no closer to melting than when you started. It would be easy to give up at that moment. But if you know the melting point of ice is 32 degrees, you know you've made significant progress. Sometimes progress is invisible. It might take some time to see the benefit of your work.
The goal isn't necessarily to change the behavior in this instant; I mean, if we get that, it's awesome, but we're looking for something else. We are looking to change how the dog feels about that event in the future.
Changing how the dog feels in a given circumstance is the key to everything. How they feel literally determines how they process information
Ralph Adolphs describes an emotion as:
"A functional state of the mind that puts your brain in a particular mode of operation adjusts your goals, directs your attention, and modifies the weights you assign to various factors in your mental calculations."
To put it another way, Dr. Lisa Feldman Barrett says, "You might think that in everyday life, the things you see and hear influence what you feel, but it's mostly the other way around: That what you feel alters your sight and hearing."
What you notice, how you feel about it, and how much importance you give those things are determined by your emotional state.
The idea that emotions are "functional states" differs from how most of us think about them. Most of us think of emotions as feelings, and that's not entirely inaccurate. But the distinction matters. How we feel can be equally described as our core affect, our emotions are more than that. Emotions are there to perform functions, such as finding safety, avoiding danger, or acquiring valuable resources. Sure, they are feelings, too, but for our purposes, it's more important to understand the way they affect behavior. Hunger, for example, is not just that empty feeling in our gut and a desire to snack on something. Hunger drives us to collect resources. Hungry people are more likely to buy things and less likely to throw things away! Hunger is a "functional state" that tells us to acquire and store resources. It doesn't exist to make our tummy rumble. It exists to push us into action.
It's not just hunger. All of our emotional states are designed to push us towards specific types of actions.
A frightened dog is more pessimistic. An excited dog is more optimistic. How the dog responds to any environmental change will be primarily determined by his emotional state. One way to think of this is that every emotion has a "palette" of behaviors related to it. The most well-known example is the "Fight/flight" response associated with fear. That's a very simplified view of how fear actually works, the truth is it's way more complicated than that. Still, it's a great illustration of how emotions really do limit choices. The stronger the emotion, the less control the dog will have over his responses.
How the dog feels the moment before the antecedent will greatly determine how it responds to it. In a sense, the outcome of an event is decided before it occurs.
If we approach behavior in a vacuum that begins with the antecedent, we ignore the most significant piece of the chain. If we have the opportunity to change how the dog feels, we can change his choices, and we can do it without relying on the promise of a reward or the threat of punishment.
So why was the solution to the dog in the shelter so apparent to myself but not to the people working with him? I can't say for sure. I have ideas, but I can't know for sure. What I am certain of is that the more time we spend dealing with how the dog feels before the behavior chain begins, the easier it will be to get a consistent and reliable response from them. Let the dog's disposition do the heavy lifting, and he will never feel like he's being trained at all. He will just make functional choices that work for everyone. The dog in the shelter, wasn't aggressive. He wasn't trying to hurt anyone.
He wasn't being bad.
He was frustrated and wanted to do something that would make his heart sing.
Fulfill that drive, teach him when to play and when not to play, and I'm betting that behavior will be under control in no time.
