Friday, January 31, 2014

Free Play vs. Rote Learning

Are We Putting too Much Pressure on Puppies to Learn Obedience too Soon?
Originally published in slightly different form at, on 11/12, 2010.

“I have spent the past few years puzzling over why dog training is no longer working that well. Today there is much more management and less reliability...” - Dr. Ian Dunbar.

Oddly enough, the reason for this downward trend in the effectiveness of dog training may be due to one of Dr. Dunbar’s own pet projects: getting as many puppies enrolled into obedience classes as early and as young as possible. In his book, Before and After Getting Your Puppy, Dunbar writes, “From the first day you get your puppy, the clock is ticking ... everything needs to be taught right away.”

Could teaching obedience skills at too early an age be one of the reasons dog training “is no longer working that well?”

In a recent article here, Dr. Peter Gray tells the story of L. P. Benezet, a superintendent of schools in Manchester, N. H., in the 1920s and 30s, who believed that teaching arithmetic to grammar school kids too early was worse than a wasted effort; it taught poor learning and reasoning.

Benezet proposed that if students were not taught any but the most practical math skills—measuring and counting—until the sixth or seventh grade, they would learn math with less effort and greater understanding. He then performed an experiment, which proved conclusively that he was right. By not teaching rote mathematical algorithms, multiplication tables, and the like, the students in the non-math classes eventually performed much better, not just in overall reasoning skills, but in math as well.

In a similar vein, fellow PT blogger and neuropsychologist Marsha Lucas says parents should wait to teach children certain cognitive skills, and focus instead on establishing an emotional connection, providing the infant with healthy feelings of attachment and belonging. 

In her blog article, “Your Baby Shouldn’t Read,” Lucas lists twelve abilities that will help kids to grow up to be well-balanced and well-balanced adults, and that she thinks should take precedence over teaching reading skills too early. Among them are, the ability to sustain attention, better management of emotions, decreased anxiety, better social relationships, greater confidence, and several others that could be applied to puppies as well as toddlers. 

What’s the best way to achieve these goals? 

I think the simplest answer is through free play.

Ian Dunbar again: “Play and especially play-fighting and play-biting during puppyhood are absolutely essential for the development of bite inhibition and a soft mouth.”

Unfortunately, that’s not what happens at most puppy obedience classes. The puppies—whose developmental needs insist on not letting them stay focused on one thing for very longare forced to do just that, and are taught, by rote repetition, how to “obey.” Then, when the pup reaches adolescence, everything he learned in puppy class has to be re-taught, almost as if he’d never learned it in the first place.

Many moderators of such classes shrug, “That’s just how it works. You always have to keep teaching obedience skills throughout the dog’s life.”

This isn’t true! In actual fact if you wait to train your dog until he’s at least 6 mos. of age, then just like Benezet’s math students, the pups will absorb their obedience lessons much quicker, and they rarely forget anything they’ve learned, as long as it’s part of a game!

There are two reasons puppies who were stars at their obedience classes “forget” their lessons when they reach adolescence. 1) Their little minds, bodies and emotions aren’t ready to learn things like the “down/stay” or how to “heel.” And 2) by the time they reach 6 mos. or so their brains have gone through a process called neural pruning, where much of what they “learned” when they were younger goes pffft! and disappears.

Evolutionary biologist Raymond Coppinger compares the cognitive differences between a neo-natal pup and an adolescent dog to the difference between a caterpillar and butterfly, as if he were talking about two different animals. He even says that the difference can be seen in how the neonatal, “sucking” skull, becomes resorbed into the adult, predatory skull. This strongly suggests that obedience skills (most of which are analogues of the predatory motor patterns of adult wolves) shouldn’t be taught to puppies until after we see a complete transformation from the puppy to the adult skull, which happens roughly around 7 - 8 mos. (which is the age when young wolves first learn to hunt). Yet most obedience schools take pups as young as 10 weeks, some even younger.

The following exchange is being passed around the internet right now. It’s an ostensible advice column, written by a kuvasz named “Blitz.”

A frustrated adolescent dog writes to Blitz with the following question. 

Dear Blitz, 

My owner and I go to obedience every week. He acts like we are going to go somewhere fun and then when we get to obedience class, I can’t wrestle with my friends or sniff or anything. What is the point? I know how to do it. I didn’t mind this when I was a puppy, but now I am six months old. What can I do? 

Signed (Pawed): Frustrated in Florida 

Here’s Blitz’s reply:

Dear Frustrated, 

You are completely missing the point of obedience class. Obedience is not supposed to be fun and games. It is an important tool to ensure that your owner does his most basic function: giving you treats. Treats are the reason that early dogs first agreed to share a cave with humans. (Treats and thunder of course.) The way to best guarantee the frequent dispersal of treats is to never respond to any of your owner’s requests too regularly. The optimum response percentage is between 30 and 60%. 

If you respond less than 30% your owner may decide that you are deaf, which will result in your visiting the vet. At the vet you may get shots and will usually have your temperature taken. Why risk it?  

If you respond more than 60% of the time, your owner will expect your response rate to increase in the future. The logical extension of that pattern is the dog who has to leap through flaming hoops to get a piece of liver. There are better ways for a dog to make a living. 

While amusing, this exchange is also reflective of the sadness Dr. Dunbar expresses when he says “dog training isn’t working that well.”

I think it’s best to follow the natural model. Remember, most obedience behaviors except the sit are analogues of the predatory motor patterns found in wolves, and wolves don’t start hunting until after adolescence. This is the model that has been set in place by Nature, and has worked for millions of years. Why change it now? Why force puppies to pay attention and “learn,” when Nature is telling them to jump around, bite, play, get distracted, and amuse the heck out of their owners?

The other problem is that it’s long been believed that a dog (or puppy) has to be calm in order to learn; dogs can’t learn when they’re highly stimulated. I’ve found that the exact opposite is true. I think it’s best to teach obedience skills as part of an active, high-energy game, where you stimulate the dog’s urge to bite, focus it on a toy, and teach him that he gets to win the toy by obeying your commands. The more actively the dog’s whole organism is involved—his emotions, his kinetic energy, his instincts, and his brain—the better and faster he’ll learn. 

This is something, that frankly, you can’t do with young puppies because they only have 3 play settings: Off, Play Hard, and Play Way Too Hard. 

I think it’s time we re-think the whole idea of puppy obedience classes, and perhaps set them up more as owner orientation classes, where the owners can watch their puppies play while the instructor explains a few simple training techniques for teaching their pup’s basic manners, but does so through the spoken and written word, without using the pup to demonstrate the process. That way the owners can learn two important things: how to teach their pup manners, at home, on their own time, and how much fun it is to watch puppies play together. That way none of our puppies will need to write to Blitz for advice. 

“Life Is an AdventureWhere Will Your Dog Take You?”
Join Me on Facebook!
Follow Me on Twitter!
Join the Rescue Dog Owners Support Group! Blog (archived)


1) “Adolescence is a period of metamorphosis - anatomical remodeling. The neonatal organism is taken apart and reconstructed into an adult. Behaviorally, the individual is remodeled from innate neonatal feeding and hazard-avoidance behaviors to the adult feeding, hazard-avoidance, and reproduction systems. Sucking feeding behaviors do not grow, or develop, into predatory feeding behaviors any more than the 18 feet of a caterpillar grow into the six legs of a butterfly. Instead, the animal is de-differentiated ... New organs are created de novo while old ones are discarded, just as the highly complex placenta and its associated behaviors are discarded at birth. Skills do not grow from the neonatal skull (the sucking skull) into an adult predatory skill. The neonatal skull is resorbed while the adult skull is being laid down.” (Coppinger, R and L, “Biologic bases of behavior of domestic dogs,” in Readings in Companion Animal Behavior, Voith, VL and Borchelt, PL, eds., Veterinary Learning Systems Co, Inc, New Jersey, 1996.)

Saturday, December 28, 2013

Dominance, Reward-Based, or Drive Training?

Which Method Makes the Most Sense?

In the Flow

There are 3 main models for training pet dogs:
1) The dominance (or pack leader) model,
2) The reward-based (or operant conditioning) model,   
3) and drive training (using a dog’s prey drive). 

The pack leader model is based primarily on stress-related behaviors found in captive wolves, and sometimes in wild wolves, but not in domesticated dogs (who don’t form packs). The positive reinforcement model originated with the ways rats and pigeons were taught to press or peck levers in a research laboratory. Natural Dog Training is based on the behaviors of a) wolves who work in harmony while hunting as a group, and b) dogs who, for thousands of years, have been trained to find and retrieve game, and to herd and guard our sheep and cattle, etc.

s My Motivation?

The motivation for learning via the dominance model comes from an instinct dogs supposedly have to submit to the authority of their “pack leader.” In operant conditioning the motive for learning is to garner “positive reinforcements” associated with obedience behaviors. Natural Dog Training describes behavior as an emotional flow system rather than as a binary, stop-and-start dynamic (dominance), or as points on a graph (positive reinforcements).

If we look at this from the dog’s point of view we can see that being dominated is not necessarily a pleasurable experience, though it might be for some dogs I suppose. Getting a reward is sometimes pleasurable, sometimes not, like when a dog is deeply involved in something that satisfies his instincts and emotions. For instance, you may be able to lure some dogs away from a squirrel using a favorite treat, but others will ignore you. Meanwhile, being in a state of flow—as happens during play—is always pleasurable. (The key to training a dog to pay attention to you instead of chasing squirrels is to provide an alternative behavior that
’s even more satisfying.)
There are other fundamental problems with both models because dominance, and reinforcers are concepts or generalized categories of things, and dogs don’t think conceptually or place things into categories. For instance, from a technical standpoint positive reinforcements aren’t actual, physical things. As mentioned before, they’re points on a graph. If they appear with enough regularity, then they qualify as reinforcements. We all know that dominance is about having a superior rank or higher status, etc., over someone else. But rank and status aren’t visceral and concrete; they’re conceptual in nature. And you can’t chase, bite or pee on a concept. That’s one problem. Another is that it’s impossible to make comparisons or to put things into categories without the ability to use and understand language. (It’s true dogs recognize verbal cues but that’s not the same thing as understanding concepts and categories.)

Each method works, or is said to work, around certain principles. Pack leader methods work via dominance and submission. Behavioral science works via punishment and reward. But drive training works via the physical properties of tension, release, and flow, meaning that dogs are motivated by a pleasurable release from internal tension or stress, thereby attaining a state of flow. It’s true that acting “submissive” or working to attain rewards may sometimes provide those feelings, which may be one reason why most trainers using these methods find that they work just fine. But only drive training is specifically designed to motivate learning in ways that dogs innately understand: “My energy was blocked but now it’s flowing!”

“My energy is flowing? What is this, dog training for hippies?” I don’t mean to offend you, but that’s utterly ridiculous!!

Your Dog, Sigmund Freud & Flow Dynamics
I’m not offended. The truth is that dogs trained the natural way are very obedient and responsive, and are very focused on their owner’s commands, as if obeying their “pack leader.” We also use a lot of food and praise. So if either a dominance trainer or a reward-based trainer saw one of us working he or she would have one of three possible reactions, depending on what we were doing or working on at the time: “This is pure dominance,” they might say, or “It looks like positive reinforcement to me.” The third response would be something along the lines of: “Wait, what did you just do? How does that work?” or “I've never seen anything like that before.”

And the thing is, when we talk about flow, we’re not being whimsical or speaking in a figurative manner. Flow is a very real and very important part of all natural systems, from tectonic plates, to ocean currents, to the way bees, butterflies and eagles ride air currents, to the way jellyfish capture prey, to the way an audience at a movie or concert are swept up in the emotions of the moment, even to the way dogs play together.

Let’s look at our own bodies. There are numerous flow systems in operation: the bloodstream, the lungs, the digestive system, the renal system, the endocrine system, even the movement of electrical impulses passing between neurons in the brain can be described as a flow system. And whenever any of these systems is obstructed in some way, it creates very real feelings of tension, pressure and stress (except in the brain, which doesn’t feel anything), and can even cause the system to shut down completely, resulting in death.

In nature, whenever a flow system meets an obstruction it puts pressure on the system, causing stress. In a river an obstruction might be a boulder. The increase in water pressure causes a smooth flow to become turbulent. If the river is obstructed by a dam, on the other hand, the water behind the dam may seem very quiet and serene; and it is. But it’s also putting a tremendous amount of pressure, structurally-speaking, on the dam. It’s only when the dam fails that you’d actually see the effects of that pressure.

In a like manner, when you become cognizant of how stress operates and exerts its influence on canine behavior, you can see the cracks begin to form before the problem has fully developed. And it’s always best to catch such problems sooner than later, if you can.

All that aside, just being alive causes emotional pressure and stress. The minute you wake up in the morning your adrenal glands increase production of a stress hormone called cortisol. The same is true for dogs. And research has shown that normal levels of cortisol motivate us to do things. In fact many of us elevate our own levels of cortisol artificially each morning with caffeine or nicotine, as a way of further motivating ourselves. And when cortisol levels go up, so does our blood pressure.
(Cortisol is also directly associated with learning.)

Cortisol was discovered in 1937. But Sigmund Freud described the way inner tension and pressure motivate behavior 17 years earlier in his treatise, “Beyond the Pleasure Principle.” He said that “the course taken by mental events … is invariably set in motion by an unpleasurable tension, and that it takes a direction such that its final outcome coincides with a lowering of that tension” resulting in “a production of pleasure.” [The Freud Reader, 594, 595.]

Freud rightly thought of the body and mind as a flow system. Internal pressures brought on by unresolved emotions or repressed instincts increase our blood pressure, just as cortisol, nicotine and caffeine do. This motivates us to behave in ways we hope might minimize those feelings. On a certain level, the same is true for dogs; they too behave in ways that they hope or have found by experience will minimize tension and stress, creating a pleasurable release.

I know it seems odd to talk about teaching a puppy to sit or heel and bring this stuff about Freud and stress hormones into the discussion. But remember, we’re discussing the differences between the ways dominance, positive reinforcement, and drive training all motivate learning. And only drive training operates specifically via the reduction of internal tension and stress.

Still, isn’t it simpler to just reward a dog? Why complicate things with all this talk about flow? The dog associates good behavior with a reward and that’s how he learns, right?

The Negative Effects of Positive Reinforcement

Unfortunately it’s not that simple. Behavior modification via positive reinforcement doesn’t work—at least not for very long—unless you constantly change the pattern of reinforcement by using one of a number of very complicated reinforcement schedules and contingencies, which if mistimed or misapplied can backfire.

Karen Pryor, a figurehead of the positive training movement, wrote on her blog in 2006: “Reinforcement may go from predictable to a little unpredictable back to predictable, as you climb, step by step, toward your ultimate goal. Sometimes a novice animal may find this very disconcerting. If two or three expected reinforcers fail to materialize, the animal may simply give up and quit on you.”

This accidentally proves (or at least strongly suggests) that behavior is not learned through positive reinforcers but through the reduction of internal tension or stress: the more stressed a dog is (up to a point)—as with the uncertainty that comes when you correctly change the pattern of reinforcement—the deeper the behavior is learned once the pattern is recognized. In fact new research shows that animals may not learn via association. They learn through pattern recognition, and that dopamine—a neurotransmitter that’s long been thought of as the brain’s “reward” chemical—is actually released when animals detect changing patterns in the environment, both pleasant and unpleasant.

B. F. Skinner formed the concepts of positive reinforcement and operant conditioning partly because he felt it was unscientific of Freud to describe behavior in terms of a person’s thoughts and feelings. How can we know with any certainty what another person or organism is thinking or feeling? Besides, even though Skinner was heavily influenced by Freud's work, quoted him extensively, and even agreed wholeheartedly with some of his major concepts, Skinner believed that all behavior was the product of conditioning, and that thoughts and feelings were irrelevant (In fact Skinner said that thoughts and feelings were products of conditioning.)

So Skinner set out to create a means of showing how conditioning takes place based on pure mathematics. However, to make sure his theory worked he fasted the animals in his research laboratory to two-thirds of their normal body weight, then placed them inside locked boxes. He did this to make sure they would only respond to environmental cues that he created, such as food suddenly appearing whenever the animal accidentally pressed a lever at the same moment that a light flashed. And sure enough the animals learned to press the lever when the light flashed, etc. But since they were subjected to physiological and psychological stress from being hungry and locked in boxes, Skinner also proved—accidentally—that stress-reduction is the real mechanism behind all learning. (Skinner would say the mechanism is irrelevant; the fact that they exhibited conditioned responses was all that mattered.)

Going back to dopamine, remember? The “reward” chemical in the brain? Freud predicted its discovery—as well as the discovery of endorphins and other, natural “feel-good” chemicals found in all mammals—in his 1930 book, Civilization and Its Discontents. Yet for years his views of psychology were deemed outmoded by many. Now, thanks to advances in neuroscience, Skinner's theories have fallen apart while Freud’s are more relevant than ever. (As one example see: “The default-mode, ego-functions and free-energy: a neurobiological account of Freudian ideas,” R. L. Carhart-Harris and K. J. Friston, Februrary 28, 2010, Brain: Oxford Journals.)

And how do dopamine and other neurotransmitters and hormones to find their way to the parts of the body and brain that need them? 
They flow.

In a 2004 article for Scientific American Mind, Dr. Mark Solms wrote: “Freud’s broad brushstroke organization of the mind is destined to play a role similar to the one Darwin’s theory of evolution served for molecular genetics—a tem­plate on which emerging details can be coherently arranged.” And: “At the deep level of mental organi­zation that Freud called the Id, the func­tional anatomy and chemistry of our brains is not much different from that of our ... house­hold pets.”

Meanwhile operant conditioning seems destined for the scientific dustbin, if it’s not there already. The only area where it’s still considered the “gold standard” by some people is in animal training: pet dogs, helper monkeys, and training dolphins and killer whales to do tricks in tanks at water parks, training that can unfortunately backfire with disastrous results, as happened at Sea World in April of 2010 when an orca killed his trainer. One has to wonder if this would have happened if the orca had been trained with the concept of flow, rather than positive reinforcement, in mind.

Okay, enough about Freud and flow and orcas and positive reinforcement. What about “pack leader” training? That seems to work just fine without all this mumbo-jumbo.

Dominance and Submission

The main problem with the “pack leader” model is that, from a scientific point of view, dominance and submission can only take place between two animals of the same species who are also members of the same social group. For instance, a wolf may be said to “dominate” other members of his pack, but not wolves from other packs, and definitely not geese, elk or human beings. By the same token, a human can’t “dominate” a dog—except in a somewhat fanciful way—because we’re members of two different species.

Another wrinkle is that pack formation is always a function of prey size, meaning canines only form packs when they need to hunt large prey such as elk or bison. For instance, wolves who live near garbage dumps don’t need to hunt large prey, so they don’t form packs. And coyotes, who are normally solitary hunters, form packs in the winter when their usual provender, small prey, is scarce. And finally, domesticated dogs don’t form packs at all, ever, not even when living a feral existence. Pack formation is always about hunting large prey, and feral dogs lack that ability. So without the capacity to form a pack how could dogs be motivated by a supposed human “pack leader?” Mind you, I’m not saying these techniques don't work, just that they don’t work because your dog thinks you’re his “pack leader.”

Still another problem is that not all packs have clearly visible dominance hierarchies. (Some scientists say these are just “latent” hierarchies.) Dr. David Mech, who many think is the world’s leading expert on wolves, once wrote that “Dominance contests are rare, if they exist at all. During my 13 summers observing the Ellesmere Island pack, I saw none.” (“Alpha status, dominance, and division of labor in wolf packs,” Canadian Journal of Zoology, 1999.)

The packs Mech studied at that time were small, mom-and-pop units (which eventually gave rise to an alternative term for the “pack leader,” the current, politically-correct term, “pack parent”). In other packs Mech studied, dominant behaviors were more prevalent, usually because of changes in pack structure, usually related to an increase in the size of the pack.

So it certainly seems to be the case that when wolves are kept in captivity, or when a pack has grown beyond the optimal size necessary for successful hunting (which peaks at around 4-5 wolves) that social friction starts to become more and more readily apparent than it is in smaller, wild packs. And, logically speaking, that’s likely to be the case because being held in captivity, or living in a pack that’s too large to sustain itself, would cause an increase in stress across the pack as a whole, bringing us back to—you guessed it—cortisol! In fact, the most “dominant” member of any animal group generally produces the highest amount of cortisol! (Remember, cortisol doesn’t cause stress, it’s just a canary in the coal mine telling us that stress is present.)

And unlike positive reinforcement techniques, whose scientific provenance is impeccable (if a bit creaky), not only is there no real scientific basis for the idea that a dogs see us as their pack leaders, there are also no scientific studies showing that dominance techniques are effective. In fact there are several showing the exact opposite.

So while dominance techniques may work with some dogs in some situations, they don’t work for the reasons “pack leader” trainers say they do.

This is why I think it’s important to understand how and why stress reduction and attaining a pleasurable state of emotional flow are the primary mechanisms for motivating all behaviors in dogs. And it’s very easy to tell when your dog is in a state of flow. He’ll have a relaxed yet dynamic demeanor, much like an athlete who’s having a good day on the ball field. But if your dog is too needy of attention, for example, it means things aren’t flowing properly. And, if he’s acting aggressive then things are flowing, but in a turbulent rather than a smooth manner.

Knowing these dynamics exist, and having the ability to see them as they play out in your dog’s daily behavior, will enable you to become your dog’s stress-relief mechanism, putting both iof you into a pleasurable state of flow, which will also put you in total control of every aspect of your dog’s behavior, but in a really cool, really fun way for both of you.

By the way, flow is the reason most dogs love to swim, chase Frisbees and tennis balls, run agility courses, go for car rides, etc. And it’s the real reason wolves follow the pack leader.” He’s able to generate more flow than they do, and they’re simply caught in his wake.

“Life Is an AdventureWhere Will Your Dog Take You?”
Join Me on Facebook!
Follow Me on Twitter!
Join the Rescue Dog Owners Support Group! Blog (archived)

Wednesday, November 27, 2013

Thoughts on Being a Dog Trainer

Why Have So Many People Suddenly Become Dog Trainers?

To me, being a dog trainer is a gift and a calling. It’s not something you can learn overnight, or become skilled at simply by hanging out at the dog run, reading books, watching Animal Planet, or by looking at training videos (though theres nothing wrong with doing any of that). Yet in the past 10 years or so the number of people in New York City who’ve decided that they’re professional dog trainers has increased exponentially. One of my clients told me that when she first arrived here with her dog, nearly everyone she met on the street or in the park claimed to be a dog trainer, and had a business card to prove it.

The truth is not everyone who claims to be a dog trainer is really equipped for the task. For instance, there’s a veterinary behaviorist I know of, who’s a widely-acclaimed expert on dogs. He gives seminars around the country. He’s written numerous books and produced several videos on training. And yet, by most reports, he’s not a very good dog trainer. He seems to have no clue as to what really makes dogs tick. On his blog, not long ago, he even asked the question: “Why isn’t dog training working as well as it used to?”

Meanwhile, I know of a guy here in the city, who makes his living as dog walker, yet in my book he’s the second-best dog trainer in New York even though no one has ever heard of him. 

Why the difference? 

Simple: one understands textbooks, the other understands dogs.

It’s also important to realize that there’s a vast difference between being an obedience instructor—someone who teaches classes on how to train dogs to sit, give paw, etc—and being a dog trainer—someone who understands all aspects of canine behavior, training and learning, and who is capable of taking any dog, at either end of the aggressive/fearful spectrum, and turning that wounded animal into a happy, emotionally-balanced, and well-behaved family pet. And in most cases, the key to becoming someone like that, someone who can do that on a regular basis, is threefold: 1) you have to have natural aptitude for it, 2) you have to spend years studying dogs, how they learn and behave, under any and all conditions, and 3) you have to either study with or have some kind of direct contact with a master trainer.

So, if you’d like to become a real dog trainer—someone who can make a real and lasting difference in a dog’s life, and in the lives of the people who love him—be aware of what’s required. 

It takes more than a business card.

“Life Is an Adventure
Where Will Your Dog Take You?”
Join Me on Facebook!
Follow Me on Twitter!
Join the Rescue Dog Owners Support Group! Blog (archived)