Originally
published in slightly different form on September 30, 2010 at PsychologyToday.com.
In my last post we learned that The Law of Effect—the idea that
animals learn new behaviors through how the consequences of those behaviors
impact the animal in either a positive or negative way—may not, in fact, be
what’s going on when learning takes place. A new model—developed by behavioral
neuroscientists—is based on identifying how and when specific dopamine neurons
in the brain do or don’t fire during a learning regimen.
This research indicates that pattern recognition may be the true agent of “reward,” meaning, specifically, that the act of paying attention to changes in the environment is what causes dopamine to be released into the brain.1
We’ve also learned that dopamine isn’t released only when something pleasurable happens to us, it’s also released when something bad happens. Yes, eating a good meal or falling in love releases dopamine, but it’s also released when we drink something noxious or get punched in the nose. It’s as if Nature is telling us, “Remember to do this!” (keep eating, keep having sex) and “Remember not to do that!” (don’t keep drinking sour milk, don’t stub your toe).
This new information about how dopamine works—by making us pay close attention to changes in environmental patterns that effect our survival and reproductive needs—means that those of us in the field of dog training now have an opportunity to better our techniques.
For example, one of the most pupular forms of dog training, the indoor training class, may be setting dogs up to fail in the real world. Yes, the puppy learns his lessons fine when he’s indoors, but get him outside, and all his learning goes out the window (so to speak). This may be part of the reason why more and more trainers are holding their classes in a local park rather than a church basement. But while this outdoorsy trend will help insure that more doggies will learn their new behaviors in a real-world setting, and that the learning may hold better outside than when it’s initially taught indoors, there’s still the problem of what’s driving the dog’s ability to learn: rewards or pattern recognition.
If it’s not rewards (which, apparently, it’s not), then what is it? If it’s pattern recognition, then some of the most basic techniques used by +R trainers need to be modified, reinvented, or re-imagined. Or do they?
There’s actually a form of training where a dog’s ability to read changing patterns in the environment—and respond accordingly—is built in. It’s how working dogs are trained.
A sheepdog wouldn’t be of much use if he didn’t respond to the patterns the flock makes when they move across a meadow. His brain comes designed with a deep-seated need to keep the flock together. So for a border collie, herding is almost a form of OCD.2
Detection dogs and tracking dogs are also trained through this faculty. Whether it’s following the scent of a perp’s trail through the Georgia swamps, locating a missing child, or a murder victim’s body, or survivors at a disaster site, or finding explosives, narcotics, or (these days) bedbugs, these dogs are highly motivated—not by external rewards—but by using their innate talent for detecting scent patterns in their environments.3
Even police dogs are trained to pay attention to patterns, i.e. whether a person is acting oddly because of some non-threatening mental or neurological defect (such as Down’s syndrome or Parkinson’s), or because that person is “up to something fishy.” It’s a mystery to me how these dogs can tune in to the difference, but they can.
Dogs are amazing animals. And I mean all dogs, not just border collies and bloodhounds and German shepherds. Every dog alive has some kind of talent that needs to be acknowledged, celebrated, and utilized in training.
Members of the positive training movement have designed and advertised themselves as the “kinder, gentler” alternative to dominance training, which is fine. It’s much better to be positive than punitive. But by defining themselves primarily as the anti-Cesar Millan, and by sticking too rigidly to the behavioral science model of dog training (which is far from “the only true model of learning” that it’s touted to be), they’re limiting their effectiveness. After all, dog owners don’t care about the science behind a training model as much as they care about whether it works for their dog.
Dr. Ian Dunbar, a key figurehead of the +R movement wrote recently, “Laboratory study has revealed a variety of reinforcement schedules. Puppy training has revealed that most of these are notorious[ly] ineffective, or impossible to administer ... Wake up! Puppy training has taught us that most of this stuff doesn’t work too well.”
Dunbar also says that one of the worst things you can do is reward a dog every time he obeys. Yet Bob Bailey, perhaps the most knowledgeable animal trainer on earth, at least when it comes to operant conditioning, says that a dog should be rewarded every single time he obeys.
Like me, Bailey has been critical of the +R movement, particularly the tendency they have to isolate themselves from the dog training community at large. Last year, after someone sent him a link to one of my articles, Bailey sent out a newsletter, writing: “The clicker training community’ has insulated itself from much of the public and from trainers not embracing the ‘purist’ methods ... How many have heard me say, ‘Beware of he or she who claims [to have] The Truth?’“
Another thing the +R training community seems to be missing is that there are really three forms of dog training: it’s no longer a choice between acting dominant/punitive or offering positive reinforcement. Before either form of training became popular there was another model, developed by Max Von Stephanitz. In fact, von Stephanitz more or less invented the idea of obedience training in the first place. His model was used primarily for training dogs for Schutzhund trials, which morphed into military and police work. But that changed in1992, when veteran police dog trainer Kevin Behan4 published Natural Dog Training, which, among other things, incorporated his knowledge of Schutzhund into a coherent model for training pet dogs as well.
Normally this wouldn’t merit much attention. After all, Behan’s model is not even close to being as well known as either the dominance or +R models. And yet, as an, I think, pretty keen observer of the changes that have taken place in dog training community during the past 18 years, I’ve found that almost every dog trainer in America is now using bits and pieces of the philosophy and even some of the techniques that Behan outlined in his 1992 book, without being aware of the provenance behind them.5
My purpose isn’t just to assign credit where credit is due (though that’s not a bad idea), it’s to point out that there are really no longer only two models of dog training. Thanks to Behan, the Von Stephanitz model has become incorporated, in one way or another, into both the dominance and +R mainstream. (I’ve even seen Cesar Millan use techniques that seem to have come directly from Behan’s 1992 book.)
As I mentioned last time, Cesar Millan has invited Ian Dunbar (and Bob Bailey) to write a chapter in his new book. This is an interesting development, especially since it coincided with my decision to write about a proposed unified dog theory. Tellingly, Millan has slowly incorporated some +R techniques into his training practice. But how many +R trainers are learning new ideas from him?
If not, why not? None of us is perfect. We can all learn from each other, as long as we’re willing to keep an open mind, and not claim that we’re the only ones with access to “the truth.”
LCK
“Life Is an Adventure—Where Will Your Dog Take You?”
Footnotes:
1) We should probably be calling these dopaminergic circuits in the brain “attentional pathways” not “reward pathways.”
2) Since some forms of OCD are based on compulsively repeating patterns—behaviors that are reinforced by the release of opiates in the brain—it’s not surprising that some border collies display OCD-like behaviors.
3) About ten years ago or so, a new type of training for detection dogs was introduced, one that was based solely on shaping behaviors through operant conditioning and external rewards. The result was dogs who would give false positives just to get a treat. Dr. Lawrence J. Myers, an expert on canine olfaction at Auburn University’s College of Veterinary Medicine, says, “Dogs want rewards, and so they will give false alerts to get them.”
4) Kevin Behan’s father, Jack Behan, was the Cesar Millan of the 1950s: he was the first “dog trainer to the stars,” the first dog trainer to have his own TV show, and the first dog trainer to run a multi-million dollar business. The Monks of New Skete actually came to Kevin Behan’s father for advice before starting their own dominance-training program.
5) The problem is that, just as most trainers don’t know about that the “stay,” the “down,” and the “heel,” etc originally came from the wolf’s hunting behaviors, neither are they aware that when they teach dogs in their puppy classes to learn the “Watch Me!” game, they’re using a technique invented by Kevin Behan. Nor do they realize that the shift many +R trainers have started making, away from a solely-food based reward system toward one that also uses play as a reinforcer, also comes from Max Von Stephanitz by way of Kevin Behan’s 1992 book.
This research indicates that pattern recognition may be the true agent of “reward,” meaning, specifically, that the act of paying attention to changes in the environment is what causes dopamine to be released into the brain.1
We’ve also learned that dopamine isn’t released only when something pleasurable happens to us, it’s also released when something bad happens. Yes, eating a good meal or falling in love releases dopamine, but it’s also released when we drink something noxious or get punched in the nose. It’s as if Nature is telling us, “Remember to do this!” (keep eating, keep having sex) and “Remember not to do that!” (don’t keep drinking sour milk, don’t stub your toe).
This new information about how dopamine works—by making us pay close attention to changes in environmental patterns that effect our survival and reproductive needs—means that those of us in the field of dog training now have an opportunity to better our techniques.
For example, one of the most pupular forms of dog training, the indoor training class, may be setting dogs up to fail in the real world. Yes, the puppy learns his lessons fine when he’s indoors, but get him outside, and all his learning goes out the window (so to speak). This may be part of the reason why more and more trainers are holding their classes in a local park rather than a church basement. But while this outdoorsy trend will help insure that more doggies will learn their new behaviors in a real-world setting, and that the learning may hold better outside than when it’s initially taught indoors, there’s still the problem of what’s driving the dog’s ability to learn: rewards or pattern recognition.
If it’s not rewards (which, apparently, it’s not), then what is it? If it’s pattern recognition, then some of the most basic techniques used by +R trainers need to be modified, reinvented, or re-imagined. Or do they?
There’s actually a form of training where a dog’s ability to read changing patterns in the environment—and respond accordingly—is built in. It’s how working dogs are trained.
A sheepdog wouldn’t be of much use if he didn’t respond to the patterns the flock makes when they move across a meadow. His brain comes designed with a deep-seated need to keep the flock together. So for a border collie, herding is almost a form of OCD.2
Detection dogs and tracking dogs are also trained through this faculty. Whether it’s following the scent of a perp’s trail through the Georgia swamps, locating a missing child, or a murder victim’s body, or survivors at a disaster site, or finding explosives, narcotics, or (these days) bedbugs, these dogs are highly motivated—not by external rewards—but by using their innate talent for detecting scent patterns in their environments.3
Even police dogs are trained to pay attention to patterns, i.e. whether a person is acting oddly because of some non-threatening mental or neurological defect (such as Down’s syndrome or Parkinson’s), or because that person is “up to something fishy.” It’s a mystery to me how these dogs can tune in to the difference, but they can.
Dogs are amazing animals. And I mean all dogs, not just border collies and bloodhounds and German shepherds. Every dog alive has some kind of talent that needs to be acknowledged, celebrated, and utilized in training.
Members of the positive training movement have designed and advertised themselves as the “kinder, gentler” alternative to dominance training, which is fine. It’s much better to be positive than punitive. But by defining themselves primarily as the anti-Cesar Millan, and by sticking too rigidly to the behavioral science model of dog training (which is far from “the only true model of learning” that it’s touted to be), they’re limiting their effectiveness. After all, dog owners don’t care about the science behind a training model as much as they care about whether it works for their dog.
Dr. Ian Dunbar, a key figurehead of the +R movement wrote recently, “Laboratory study has revealed a variety of reinforcement schedules. Puppy training has revealed that most of these are notorious[ly] ineffective, or impossible to administer ... Wake up! Puppy training has taught us that most of this stuff doesn’t work too well.”
Dunbar also says that one of the worst things you can do is reward a dog every time he obeys. Yet Bob Bailey, perhaps the most knowledgeable animal trainer on earth, at least when it comes to operant conditioning, says that a dog should be rewarded every single time he obeys.
Like me, Bailey has been critical of the +R movement, particularly the tendency they have to isolate themselves from the dog training community at large. Last year, after someone sent him a link to one of my articles, Bailey sent out a newsletter, writing: “The clicker training community’ has insulated itself from much of the public and from trainers not embracing the ‘purist’ methods ... How many have heard me say, ‘Beware of he or she who claims [to have] The Truth?’“
Another thing the +R training community seems to be missing is that there are really three forms of dog training: it’s no longer a choice between acting dominant/punitive or offering positive reinforcement. Before either form of training became popular there was another model, developed by Max Von Stephanitz. In fact, von Stephanitz more or less invented the idea of obedience training in the first place. His model was used primarily for training dogs for Schutzhund trials, which morphed into military and police work. But that changed in1992, when veteran police dog trainer Kevin Behan4 published Natural Dog Training, which, among other things, incorporated his knowledge of Schutzhund into a coherent model for training pet dogs as well.
Normally this wouldn’t merit much attention. After all, Behan’s model is not even close to being as well known as either the dominance or +R models. And yet, as an, I think, pretty keen observer of the changes that have taken place in dog training community during the past 18 years, I’ve found that almost every dog trainer in America is now using bits and pieces of the philosophy and even some of the techniques that Behan outlined in his 1992 book, without being aware of the provenance behind them.5
My purpose isn’t just to assign credit where credit is due (though that’s not a bad idea), it’s to point out that there are really no longer only two models of dog training. Thanks to Behan, the Von Stephanitz model has become incorporated, in one way or another, into both the dominance and +R mainstream. (I’ve even seen Cesar Millan use techniques that seem to have come directly from Behan’s 1992 book.)
As I mentioned last time, Cesar Millan has invited Ian Dunbar (and Bob Bailey) to write a chapter in his new book. This is an interesting development, especially since it coincided with my decision to write about a proposed unified dog theory. Tellingly, Millan has slowly incorporated some +R techniques into his training practice. But how many +R trainers are learning new ideas from him?
If not, why not? None of us is perfect. We can all learn from each other, as long as we’re willing to keep an open mind, and not claim that we’re the only ones with access to “the truth.”
LCK
“Life Is an Adventure—Where Will Your Dog Take You?”
Footnotes:
1) We should probably be calling these dopaminergic circuits in the brain “attentional pathways” not “reward pathways.”
2) Since some forms of OCD are based on compulsively repeating patterns—behaviors that are reinforced by the release of opiates in the brain—it’s not surprising that some border collies display OCD-like behaviors.
3) About ten years ago or so, a new type of training for detection dogs was introduced, one that was based solely on shaping behaviors through operant conditioning and external rewards. The result was dogs who would give false positives just to get a treat. Dr. Lawrence J. Myers, an expert on canine olfaction at Auburn University’s College of Veterinary Medicine, says, “Dogs want rewards, and so they will give false alerts to get them.”
4) Kevin Behan’s father, Jack Behan, was the Cesar Millan of the 1950s: he was the first “dog trainer to the stars,” the first dog trainer to have his own TV show, and the first dog trainer to run a multi-million dollar business. The Monks of New Skete actually came to Kevin Behan’s father for advice before starting their own dominance-training program.
5) The problem is that, just as most trainers don’t know about that the “stay,” the “down,” and the “heel,” etc originally came from the wolf’s hunting behaviors, neither are they aware that when they teach dogs in their puppy classes to learn the “Watch Me!” game, they’re using a technique invented by Kevin Behan. Nor do they realize that the shift many +R trainers have started making, away from a solely-food based reward system toward one that also uses play as a reinforcer, also comes from Max Von Stephanitz by way of Kevin Behan’s 1992 book.
.
ReplyDelete