Mice & Mutts: Why Behavioral Science Is Losing the Training Wars

Dogs: Domesticated Arctic Gray Wolves or Sophisticated Norwegian Brown Rats?

Originally published in slightly different form on October 9, 2009 at PsychologyToday.com.

You may not be aware of it, but there’s a quiet war raging right now in the dog-training world. It’s a conflict between positive reinforcement (+R) trainers and behaviorists like Ian Dunbar and Nicholas Dodman who base their methods on the principles of learning theory. They’ve pitted themselves against traditional or dominance trainers like Cesar Millan and the Monks of New Skete, who follow the alpha theory.

Learning theory: whenever a behavior is followed by a positive consequence (i.e., a reward), it will tend to become learned. Negative consequences (or punishment), will tend to extinguish unwanted behaviors. These principles are said to be true of all learning in all animals, from rats to monkeys to humans. It’s clinical, it’s clean, it’s almost mathematical, and a great deal of modern cognitive science is based on its principles.

The alpha theory: dogs have an instinct to obey anyone they view as their “pack leader.” When a dog disobeys or refuses to learn, he’s dominant. When he does obey, he’s submissive. This theory is based on how members of a wolf pack supposedly follow the alpha wolf, and only applies to domesticated dogs; you can’t train lab rats and helper monkeys or wild wolves via dominance.

In terms of results, each side is equally successful, yet one essentially treats dogs like lab rats while the other sees them as wolves. Neither seems to treat dogs as, well ... dogs.

The war actually started in the 1930s, pitting American psychologist B. F. Skinner against Austrian (and Nazi) ethologist Konrad Lorenz¹. It heated up again in 1993 when Ian Dunbar founded the Association for Pet Dog Trainers (APDT) to try to steer dog training away from the practices of traditionalists like the Monks of New Skete who at the time asked, “How hard should you hit your dog? If she doesn’t yelp in pain, you haven’t hit her hard enough.” Dunbar and the APDT thought this wasn’t the proper way to treat your best friend, and wanted to change the way trainers treated the animals in their care. (I’m on Dunbar’s side here, at least in this one regard.)

The APDT quickly grew to become the world’s largest organization dedicated to the training of pet dogs. Since its inception there has been a worldwide explosion of puppy classes run by reward-based trainers. By the beginning of the 21st Century it seemed as if Dunbar’s utopian ideal for the humane training of all dogs was about to be realized.

But in 2004 the National Geographic Channel premiered their reality show, The Dog Whisperer, starring, as his critics liked to characterize him, “an ex-dog groomer,” named Cesar Millan. Nicholas Dodman, of Tufts University, sent a letter of protest to NGC, stating that he and his colleagues believed Millan had “set dog training back by 20 years.” The American Humane Society urged National Geographic to take Millan off the air, expressing dismay at the “numerous inhumane training techniques” he uses, including hanging dogs from a choke chain until they lose consciousness. In an article in Esquire, “The Dog Whisperer Should Just Shut Up,” Patricia McConnell, a zoologist from the University of Wisconsin, joined in. “Behavioral problems are the result of miscommunication,” McConnell said. “Either dogs don’t know what their owners want or humans inadvertently teach them to do the wrong things. Most behavioral problems can be solved using the science of how animals learn.”

You might be thinking, “So what? Who cares about this petty professional jealousy and squabbling? It’s pretty meaningless in the overall scheme of things, isn’t it?”

Perhaps. But science writer Virginia Morell wrote in a recent article in Science Magazine, “Dogs are fast becoming the ‘it’ animal for evolutionary cognition research. Our canine pals, researchers say, are excellent subjects for studying the building blocks underlying mental abilities, particularly those involving social cognition.”

So this conflict in the dog training world might be able to reveal something important about the true nature of dogs in particular, and animal consciousness in general. After all, how can two diametrically-opposed scientific theories be equally successful? Oddly enough, while both theories get results a fairly large percentage of the time, dogs can also be trained without using any aspect of either theory. For instance, you can sometimes increase a dog’s obedience tenfold by acting submissive!² It’s also possible to completely extinguish some behaviors in dogs simply by rewarding the dog every time he does the very thing you don’t want him to!³ You can’t train rats and monkeys in such backwards fashion, but for some reason it works with dogs.

We now know (though Cesar Millan doesn’t seem to) that in wild wolf packs there is no pack leader, at least not in the traditional sense of the phrase,⁴ and that in most wild packs dominant and submissive behaviors are so rare as to be virtually nonexistent. Also, none of the behaviors Millan uses to convince dogs he’s their pack leader actually exist in nature. In other words, there is no scientific foundation for his form of training. But if Dunbar and Dodman truly have science on their side, as they believe, then why haven’t they dominated the training landscape? (Ironically, Dodman himself has written a carefully detailed account of how ineffective operant conditioning is at solving behavioral problems in dogs, although that was clearly not his intention.)⁵

Dogs are able to detect diseases like cancer and diabetes, they can predict epileptic seizures, find escaped convicts, missing children, they can even find their way home after their owners have moved to a new city or new county miles away. It shouldn’t be surprising then that they can detect the holes, flaws, and cracks in our theories about them and their behaviors, fetch those flaws for us, and drop them at our feet. And what they’re showing us, in these silly training wars of ours, the thing they’ve detected that our cleverest scientists have overlooked, is that there’s a (perhaps) tiny piece of the smooth-running operant conditioning machinery that’s simply inoperative.

True, the alpha theory has very little basis in science. But it’s starting to look like learning theory isn’t as scientific as we’ve all been led to believe it is either. Yet it’s so deeply embedded into our scientific studies and academic institutions that it’s nearly impossible for us to view it as incomplete and inaccurate, let alone divest ourselves of its hold over our ability to see how learning really does take place. But dogs are innocents. They’re like the child in the Hans Christian Anderson fable. And they’re telling us that on a certain level operant conditioning, seemingly arrayed in the purple robes of science, isn’t wearing any clothes.

That’s why behavioral science is losing the training wars.

Next time I’ll explain how behavioral science is not only failing our dogs but our kids as well, and why the two are connected. I’ll also offer a simple theory to explain why dominance training works with some dogs, why operant conditioning is ineffective at solving behavioral problems, and why dogs can be taught without punishment or rewards.

LCK 

“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!

Footnotes:

1) Science researcher Boria Sax unearthed evidence showing that Lorenz was eager to join the Nazi party. And his job under Hitler reportedly involved determining which offspring of mixed Polish and German heritage had enough Aryan blood to remain in the gene pool and which didn’t and had to be sterilized or exterminated.

2) Years ago I did an experiment to test the alpha theory. I put myself in what’s presumed to be a submissive position (rolling over on my back) to a number of dogs (one at a time, in various settings and situations) as part of a game, and found that all the dogs were more obedient after I’d “acted submissive” than they had been before. (To learn more, read my personal blog article, “The Proper Way to Do an Alpha Roll.”)

3) I also did an experiment where I verbally praised my dog every time he decided to scavenge something from the streets of New York, chicken bones, pizza crusts, you name it. Within three days he stopped all scavenging, and never showed any interest in it again. In other words, I successfully extinguished a behavior simply by rewarding the dog every time he exhibited it. From the perspective of learning theory, his scavenging was always followed by a positive consequence, and yet instead of reinforcing that behavior, it somehow eliminated it. (Read my personal blog article, “Praise to Correct.”)

4) See my blog article: “Pack Leader or Predator.” Also, dominance—if it exists—can only take place between animals who are members of the same species and the same social group. This means that no human can act dominant toward a dog; they’re two different species. And no trainer can effectively train a family dog via dominance because he’s not part of the dog’s social group/family.

5) The Dog Who Loved too Much