Monday, January 18, 2010

Siege at Druid Peak

This article also appears at

Wolves, Social Networks, & Feedback Mechanisms
"Nothing in nature is random." - Spinoza

This is one of the strangest and most intriguing stories I've ever come across. It starts simply enough with a pack of gray wolves living happily in British Columbia. Then one day, in 1995, while they were out doing ordinary wolflike things, they were tranquilized by a group of biologists, fitted with radio collars, then transported to a new environment: the Lamar Valley in Yellowstone National Park. They were dubbed the Druid Peak pack, named after a central geographic feature of the valley. By 2002 -- after six or seven generations of new wolves came along -- the pack was getting a little too big to sustain itself. So a group split off, left the valley, and formed their own pack near Slough Creek, where, over the next several years, they grew to outnumber their old pack mates by almost three to one.

In 2005, when PBS began filming In the Valley of the Wolves, to document this phase of the wolves' transition to the park, the Druids had been ensconced in the Lamar Valley for over twelve years. They were reportedly now "at war" with the Slough Creek Pack, though the incursions from their rivals were few and far between. There was also a coyote husband and wife living in the valley, enjoying a semi-peaceful co-existence with the Druids. They would often approach the Druids' latest kill from a safe distance, and, in an almost pro forma way, one or more of the Druids would launch a mock attack.

Wolf: "Hey! You know the rules!"

Coyote: "Sorry, we were hungry. We'll come back later."

Wolf: "Okay. But there might not be much left..."

Even when the Druids were hunting it was almost like a game to them. It was only when they got in close enough to be gored or maimed by their prey's horns and hooves that their teeth came out.

There was also a lone wolf who apparently wanted to leave the Slough Creek pack. He would occasionally come around, trotting behind the Druids at a safe distance, eyeing a particular young female. The father repeatedly chased him away, not in a mean fashion, just a kind of, "She's too young for you!" The female's attraction was stronger than her father's objections, though, so her suitor was eventually allowed to join the family.

And family is the key word. If you were to compare life in the Lamar Valley to a 1950s television show, it would've been more like "Leave it to Beaver" than "Wild Kingdom."

That all changed in the winter of 2006 when the Slough Creek pack came into the valley, launching an all-out attack on everyone in it. In a period of just a few days they had killed the mama and papa wolf, scattered the rest of the pack, slaughtered more elk than they could eat, and instead of just chasing the coyote couple away from their latest kill, they systematically chased down the husband and ripped him to shreds while his helpless, now-pregnant wife watched, terrified, from a distance. Then, their thirst for blood still not satisfied, they came after her too. Luckily, she was able to scramble safely away.

Then spring came slowly, as it does in the high country. Several of the Slough Creek females had given birth, and were raising their litters in a group of dens on the side of a hill sheltered from the lingering snow. The valley's victors, the Slough Creek males, had grown a bit lazy. They had let their guard down. So they weren't prepared when, one day, as they returned from a hunting expedition, they found a group of mysterious black wolves -- very different in color from the mostly grey and brown Druids and Slough Creekers -- who had come marching into the valley, taken strategic positions on the hill, and staged a siege outside the dens of the nursing females. These interlopers -- while fewer in number than the Slough Creek males -- easily fended off all attempted sorties. There was no way for those males to get food to their wives and pups.

The black wolves did nothing but wait patiently on the side of that snow-covered hill. The days turned into weeks, and one by one, every single Slough Creek pup in every litter died slowly of starvation. The helpless adults were bereft, agitated, in a state of terrible distress, so much so that by the time the black wolves finally left the valley, never to return, the Slough Creek pack scattered to the winds, their spirits broken.

It wasn't long before the Druids came out of hiding, joyously re-assembled, and re-took control of their beloved valley. For them, it must have been the best spring in years.

As mysterious as this event was, on a certain level we can sort of understand the behavior of the Slough Creek pack. They wanted the valley. They had superior numbers; they came in and they took it. Still, they didn't have to kill the coyote husband; he was no threat. In fact, by chasing him down and killing him, they used up more energy than it was worth. So did killing more elk than they needed. If one of the laws of nature is the conservation of energy, then these wolves weren't aware of it.

Even so, that's not what's really puzzling. What's really hard for us to wrap our minds around is the behavior of those mysterious black wolves. They weren't following the rules of nature either. Far from it! They weren't after the valley's resources. They didn't do much if any hunting there. They just waltzed in, surrounded the dens, waited for all the pups to die, and then left, as if that were their sole purpose, which makes no sense at all. Who were they? Where did they come from? Why did they do what they did?

Biologists and evolutionary scientists can't explain this. To them it's either an extended example of biological altruism (their behavior benefited the Druids), or just a random event that holds little or no meaning.

Yet we also have no explanation (other than, "Gee, aren't dogs wonderful?") for incidents like the one reported on NPR last week: A pointer named Effie, was out for a normal walk with her owner. But within a few minutes she started pulling to go in a different direction, than took off running to a nearby house where a 94-year-old man was lying unconscious, face down in his driveway. The dog started licking the stranger's face. Her owner called 911, then began doing CPR. Together, they saved the man's life.

Why do these things happen? How do we explain them?

The only answer I can think of is that that dogs and wolves may have their own form of Twitter and Facebook, their own social networks that help them tune into situations that require action when someone in the network is in danger, or perhaps even when the network-as-a-whole is out of whack. The difference is that we react to situations like the one going on right now in Haiti, or the Tsunami several years ago, both from a gut level, as animals do, and from a plane of conscious thought. "Oh, those poor people!" we think, and wonder how we can help. Then we start networking.

Dogs and wolves can't text each other; they don't have thumbs. But they can definitely feel what someone else is feeling, and they seem to do so as if it's actually happening to them. From my observations, canines also seem to have a gut reaction when something's not right in whatever social network, large or small, they're a part of. For Effie, that network may be her neighborhood. For those mysterious black wolves, it's a much larger network, one that includes the entire Yellowstone basin. In fact I would argue that it consists of all wolves and coyotes, along with the birds, the elk, the aspen trees, the rivers and streams, the weather systems, the sky above and the earth below, and even the PBS cameraman, drinking coffee from a Thermos, munching an energy bar, and making breath clouds behind his telephoto lens while waiting for the wolves to do something interesting.

Konrad Lorenz, who's responsible for most of the misinformation we now have about dogs and wolves, was still a brilliant scientist, capable of keen insights. One of his theories on ecology is the idea of feedback mechanisms. "In nature," he writes, "these mechanisms tend towards a 'stable state' among the living beings of an ecology: A closer examination shows that these beings... not only do not damage each other, but often constitute a community of interests. It is obvious that the predator is strongly interested in the survival of that species, animal or vegetable, which constitutes its prey. ... It is not uncommon that the prey species derives specific benefits from its interaction with the predator species." (Civilized Man's Eight Deadly Sins, 1974, p. 33). As if to prove this point, after the Canadian wolves were relocated to Yellowstone, a funny thing happened; the aspen trees, which had been dying out, made a comeback.

I don't think the black wolves came to Yellowstone to rescue the Druids from the evil Slough Creek pack. It's easy for us to see things in those terms, black and white, heroes and villains. But I think the black wolves came simply because they were part of a natural Lorenzian feedback mechanism. The Slough Creek pack had gotten too big and powerful for the network-as-a-whole. The black wolves felt a disturbance in that network, so they took action, not to punish or "unfriend" the Slough Creek wolves, and not to rescue the Druids either, but simply to restore the network to its optimal setting. (It's telling that they didn't kill any of the adult wolves; they simply got in the way of their ability to provide food to their young, which prevented a new generation from becoming part of a pack that had already grown too big.) To me this is the only explanation that makes any sense.

The key feature of wolf behavior is their ability to hunt together as a cohesive social unit. In order to do that, you have to have enormous social and emotional flexibility. You have to be able to read each other's social signals very quickly and extremely well, especially in a high-pressure situation. When dogs begans domesticating us, around 12,000 years ago, they expanded on the wolf's natural social networking skills, so that now they exhibit an extraordinary ability to read human social signals in a way that's far more developed than what we find even in our closest biological relative, the chimp. (In some ways dogs are far more social than even we are!) Their social intelligence is the primary reason dogs are the current "it" animal for cognitive research.

So the next time you take your dog for a walk, switch off your cell phone or Blackberry. And don't worry that you'll be out of touch; you'll still have a direct connection to one of the most wonderful and miraculous social networks ever created, right there at the end of the leash.

Monday, January 11, 2010

Why Do Dogs Mark Their Territory?

Why Do Dogs Mark Their Territory?

The simplest answer is they don't.

In The Intelligence of Dogs, author Stanley Coren gives us the classical explanation of this myth: "All canids use urine ... to mark the limits of their territories. In males this marking behavior is usually accompanied by leg lifting to direct the urine against large objects (trees, rocks, bushes) to place the scent at nose height for other dogs and to allow the scent to radiate over a large area. Some African wild dogs ... scrabble as high up the trunk of a tree as possible before squirting their message."

First of all, dogs urinate far beyond the boundaries or limits of their so-called territory. Secondly, males aren't the only ones who lift their legs; some females (usually the anxious type) do this as well. Thirdly, dogs don't just urinate on large objects, but on vertical objects (trees, posts), unfamiliar or inorganic objects (tires, plastic bags, fire hydrants), and on anything carrying a scent that the dog wants to cover (such as another animal's urine or feces).

Coren is not responsible for the myth. His offhand re-telling of it, as if it were a scientific certainty, merely highlights a general tendency in science: in attempting to dissect how an animal's behavior might serve an adaptive purpose -- in this case marking would be a hypothetical means of limiting competition within a niche or habitat -- most scientists blur the line between what makes sense in terms of the grand arc of evolution, and what an individual animal is capable of in terms of its cognitive abilities.

Of course if a biologist who witnessed the African wild dogs madly scrabbling up the tree trunks did so with the belief that dogs urinate to send a message to other dogs, their behavior would, no doubt, confirm his hypothesis. But if we approach this behavior with a clearer mind we might ask, how could these dogs possibly know the "nose height" of another, purely hypothetical dog who might (or might not) come along at some undetermined point in the future?

In The Secret Life of Dogs, anthropologist Elizabeth Marshall Thomas compares a dog's urine marks to Hansel and Gretel's trail of pebbles and bread crumbs: a means of finding their way home. But that would mean the dog was planning for the potential (thus, hypothetical) possibility of getting lost. I've never seen any evidence for that kind of thinking in the dogs I've known. They live totally in the now moment.

Roger Caras, whose voice used to be heard each year at the Westminster Kennel Club Dog Show in New York, was fond of saying that when a dog sniffs a fire hydrant he's "reading his mail". This is highly anthropomorphic, yet it's hard to dispute that a dog does get information from the scent of other dogs this way. The only question is, did the dog who left the scent do so with the intention of "sending a message" to the one who comes along later?

It doesn't make sense that any dog would have the intelligence necessary to leave messages for other dogs in this manner because in order to do so he would have to be capable of propositional and/or hypothetical thinking, directed fantasy, mental time travel, not to mention a full-blown theory of mind. "If I mark this fence (propositional thinking), Spike will come along some time in the future (directed fantasy, hypothetical thinking, mental time travel), sniff it (more fantasy, more hypothetical thinking), and know he's in my territory (theory of mind, abstract and conceptual thinking) and start to feel nervous about being here (more theory of mind)." That's pretty complicated thinking for a dog.

Territory is defined by biologists as an area which an animal will defend against intruders of the same species. But how is such an area delineated in the animal's mind? Are its boundaries visible and concrete or imaginary and abstract? Is a dog capable of forming a mental image of where his territory begins and ends? And if animals have no sense of self and other, how could they think of a territory as "mine" or "belonging to me?"

Meanwhile, I got my first glimpse into a more reasonable explanation of why dogs mark -- one that has nothing to do with "territory" -- many years ago when I took my dog Freddie to a training session I had with a six-month Maltese male named Buckwheat who hadn't had much socialization with other dogs. Freddie's presence made Bucky a little nervous, but not to the point that he couldn't learn the games we were teaching him. However, at several key points during the lesson -- which was taking place in the dining room -- I put Freddie in a down /stay by a piano in the living room, to keep him out of the way. Later Bucky's owner told me that immediately after Fred and I left, Bucky had gone over to the piano and had urinated on the spot where Freddie was told to lie down.

Why did he do that? To mark the limits of his territory? He was already inside his territory; in fact, not even inside his territory but inside his "den." Did he do it to send a message to Freddie? The answer is simple: The rug held remnants of Freddie's scent. That made Bucky nervous so he put his own scent on top, to cover it up. Yes, in a sense, he marked the piano, but not to tell Freddie that it was his. If anything, he did it just to relieve his internal tension.

A few years later, while doing research for a subplot about kidney disease for my 4th novel, 'Twas the Bite Before Christmas, I learned that in mammals, the need to urinate is controlled, in large part, by the neuropeptide vasopressin. Higher levels of vasopressin increase water retention, reducing the need to urinate. Low levels are associated with excessive urination, bedwetting, etc. Vasopressin also has a converse relationship with the stress hormone, cortisol: when cortisol levels go up, vasopressin goes down, suggesting that there's a causal relation between stress and excessive urination.

So it seems far more likely that when one male detects the scent of another dog, (particularly an unknown male), it could cause a perhaps low-level stress reaction, which would then increase his need to urinate. As he does he would feel the pleasure of releasing some of the tension and pressure  in his body. And thus, over time, his body would self-reinforce the behavior of peeing on top of another dog's scent. It would be a purely emotional and perhaps Pavlovian response, not based on intellect or other mental faculties.

Later, when this dog smells a urine mark he'd made earlier, he would probably re-experience, on some level, the original lessening of tension and the pleasure it produced. (Of the five senses the sense of smell is the one most likely to evoke memories.) He learns to mark in order to relieve emotional tension and to feel connected to his environment.

This explanation is simple, whole, and complete. It requires no complicated thinking on the dog's part. It obeys the rules of parsimony and logic. And it only requires that a dog have the ability to experience tension and pleasure, and to form simple physical and emotional associations.

Sunday, January 3, 2010

The Natural Dog Training Difference

Here's an in-depth look at how Natural Dog Training differs from the popular pack leader and positive reinforcement methods, including videos of all three methods in action.

The Natural Dog Training Difference
I read something interesting on Kevin Behans’s blog the other day, about how nearly everyone who takes their first ride on a camel or elephant experiences motion sickness, but this doesn’t happen when people ride a horse for the first time. Kevin’s reasoning is that horses naturally know how to adjust their movements to incorporate the rider’s center of gravity.

In thinking about that, I realized that what’s missing from both the dominance and +R approach to training, and what we do, is that dogs really do have a sort of emotional center of gravity as Kevin postulates. And when we teach them to do an exercise like the heel, for instance, using thought-centric models of learning, such as dominance and +R, the dogs have to figure out, on their own, how to match their forward momentum and energy with ours. But when we teach them using Natural Dog Training, no matter how bad we are at it initially, if our goal is to teach the dog to be in-synch with us physically and emotionally (instead of teaching them to respect our leadership, or by rewarding their external behaviors), at some point we’ll find that we’re actually creating a feeling in the dog of a shared center-of-gravity, just like with a horse and rider. In that respect, heeling not only feels natural to the dog. It feels really, really good. 

To highlight these differences, here’s a video of Cesar Millan [note: this video has been deleted] solving a fairly simple behavioral problem of a great Dane who gets too energized when she goes jogging with her owner. She expresses this excess energy (which is essentially a nervousness about how to keep her desire to run full-bore in check) as jumping up. Millan interprets this as dominance, and teaches the dog to stay next to her owner by being “submissive.”

There’s so much wrong in Cesar’s explanation of the problem, not mention in how he solves it, yet it’s hard to dispute the visual “evidence” of the results.

So what, exactly, if anything, is wrong with Cesar’s approach? 

Well for starters, even without knowing that dominance is not a real character trait or behavioral output in dogs, it’s quite easy to see that the Dane is simply expressing a strong social attraction — the desire to connect to her owner — in an inappropriate way. She’s also feeling nervous because she's unsure of how to align her energy with the owner while they're running: faster forward momentum = a challenge to the dog's ability to feel or stay connected, so she’s jumping up to ground some of that excess energy. 

The thing is, you never want to punish or correct social attraction, no matter what form it takes. And you especially don’t want to do it by intimidating the dog, as Cesar does. If you watch his body language, and the body language of the dog very closely, you’ll see that Millan is actually acting very much like a predator in order to keep the dog in a “submissive” state. You can see this most clearly in the section where he first demonstrates the “touch” to the dog’s throat — which in the past he described as a “bite,” as in “If a dog can bite me, why I can't I bite him back?” — then moves into the dog’s space, making the dog even more nervous and unsure of herself.

In short, the problem is “solved” by repressing the dog’s energy instead of celebrating that energy and channeling it into a happy, joyous heel. (Personally, I probably wouldn’t take a great Dane jogging anyway; I don’t thinking jogging is a good idea with most dogs, particularly those with a big barrel chest and narrow waist.*)

But other than telling the owner to come up with an alternative exercise plan,** if we look at this as an energy problem, the way to solve it would be keep that level of energy active in the dog, but give it a different outlet without intimidating or repressing her drive to connect. In other words, keep the dog’s drive energy up but channel it into a heel. (After a while the Dane learns to do this on her own, but is still confused, unhappy, and not as energetic as before.)

Contrast the dominance approach with the traditional clicker-training and food-luring method, as shown by Nancy Cusick, a professional dog trainer from Texas who's been described (by herself and others) as The Awesomest Dog Trainer in Austin, which she may very well be. (I pulled this video at random from YouTube.)

I see several things lacking here. One is that the puppy is a bit too young for the exercise. She just wants to sniff and explore. Each time she does, Cusick redirects her with a kissing sound. That’s nice, and fine in theory, but by doing this Cusick slowly and inevitably becomes an obstacle to the puppy’s desires, desires that are being controlled more by the puppy’s developmental needs than by hunger.

Also, at one point when the pup sits while not in the heel position, the trainer moves her body next to the pup’s rather than using her own body language and energy to induce the puppy to move toward her and then sit. Then she clicks and rewards the dog for being in that position. This is based on the somewhat questionable idea that dogs learn through positive reinforcement: if the dog is reinforced while it’s in the proper position it will gradually learn to choose that position on its own. (Notice that despite the seeming validity of this idea, the more the trainer rewards the puppy for being in the heel position, the more the puppy actually wanders off to explore, and do other things on her own.)

Another problem is that when the trainer accidentally drops food on the ground, and the puppy goes after it, the trainer makes the kissing sound again to try to redirect the puppy’s attention back to her. Again, you can see clearly that the more the trainer does this, the less attention the puppy pays to the trainer. (At one point the trainer even jokes to the camera, “Attention doggie deficit…” and chuckles.) 

There is nothing inherently wrong with using a kissing sound while teaching a dog to walk next to you. The problem here is with the timing. Instead of making the kissing sound as soon as the pup loses focus, the trainer does it after the puppy has already projected its energy onto something else. So the kissing sound ends up feeling like a punishment to the puppy.  

Puppy loses focus ... finds something to focus on ... handler makes kissing sound.  

Puppy feels, “Hey, I was having fun!” 

Contrast that with making the kissing sound the instant the pup loses focus, before she finds something else to focus her energy on: “What can I find around here to focus on?”   

Puppy loses focus ... trainer immediately makes kissing sound.  

Here the puppy feels, “Oh, good! I can focus my energy on you! This feels great!”

See the difference?

I’ll give Cusick the benefit of the doubt (as I said, she probably is the awesomest dog trainer in Austin, Texas), and suggest that part of the problem may be she’s not just focused on training the pup, she’s also talking to the camera as she works: not an easy thing to do. 

However, in the end the puppy only has a “generalized” heel, whose focus is very easily broken except when doing the sit while in the heel position. The reason the puppy is focused then is because that’s the only time the puppy isn’t feeling a disconnect between its own body and the trainer’s. While they’re doing the heel the puppy is mildly interested in getting the treats, but can’t figure out how to match her body’s need for forward momentum with the movement of the trainer’s body and the food lure. And the trainer isn’t using the food to help the pup solve the problem, she’s only using it as a lure and a positive reinforcement.

To recap, in Cesar Millan’s mind the dog’s problem is “How can I be submissive to my pack leader?” which is based on a false premise. Meanwhile, the positive trainer sees the puppy’s problem as, “How can I get a reward? Maybe if I heel I’ll get a treat?” which is just as false. Both ideas require the dog to engage in a linear, rational, time-dependent thought process, and a) dogs aren't capable of rational or hypothetical thinking, and b) they live totally in the moment, without any awareness of linear, chronological time. 

In each case the real problem for the doggie is, “How can I get my body to feel in-synch with my handler’s energy and momentum while we’re both moving together?” 

Now contrast these two approaches with this video of Kevin Behan, working on the heel with a Doberman pinscher named Laszlo, using the natural approach. 

First of all Laszlo is no ordinary dog. His owner brought him to Kevin because she was having a great deal of difficulty with his overabundant energy. As she wrote on her own blog, Laszlo was so wired that “he wouldn't lie down. I don't mean on command — I mean, he wouldn't lie down. Such was his anxiety and vigilance.” 

Now, that’s a tense doggie! 

The first thing Kevin does with Laszlo in this video is the pushing exercise, where he gets Laszlo to push for food. He does it, among other things, to stimulate Laszlo's social attraction to him. To the uninitiated viewer this may look just like “luring” the dog with treats the way Nancy Cusick does, but there’s a lot more to it than that. 

How to Do the Pushing Exercise 

How and Why It Works 

After a bit of pushing, Kevin begins moving around, encouraging Laszlo to move with him. At one point Laszlo gets distracted by a puddle, but Kevin just keeps moving (no kissing sound), encouraging the dog to connect to him (and what’s in his bait bag). At other times Laszlo finds bits of food on the ground and Kevin waits a bit for him to finish eating them before he starts moving again. 

Once he’s got Laszlo moving with him he begins to oscillate between acting like prey and predator, moves that again, to the uninitiated, might seem to have no purpose.

“He’s just throwing in some silly tai-chi moves to impress people.”

They may seem silly, but if you watch carefully, you’ll see that each shift in Kevin’s body language creates an immediate, in-the-moment shift in Laszlo's behavior, his approach to staying in-synch with Kevin’s movements. Those “silly” moves of Kevin’s have nothing to do with tai-chi, per se, though they do create shifts in Laszlo’s energy (which, for all I know, may actually be one of the goals of tai-chi).

At a certain point, Kevin even gets Laszlo to hold a down/stay without even giving the command. (Kevin exhibits some interesting “marching band” moves during this sequence as a means of both enticing Laszlo to break the stay, and to keep him in it at the same time.) And you’ll notice that Laszlo’s ears never go down or back except once or twice, for a fraction of a second, and each time they do, Kevin compensates with food or with his body language to bring the dog’s energy back into to a more relaxed and confident state.

There is no intimidation or dominance and submission in anything Kevin does. (Kevin does occasionally touch Laszlo’s neck with the back of his hand, which is done to help ground the dog's energy a little and to “steer” him a little, the way you’d do with a horse’s reins.)

You’ll also notice (I hope) the way Kevin does the about turns, which gently induce Laszlo to stay in "the pocket." To help with this, he uses food as a means of keeping the dog’s drive-to-connect up and active rather than as a reward for any one specific behavior. It’s more like a dance, one that leaves Laszlo entirely under Kevin’s command with no punishment or bad feelings taking place. 

At the end, Laszlo is heeling off-lead, and his energy is totally plugged in to Kevin.

By the way, the movements that Kevin makes don’t need to be, and in my opinion, shouldn’t be, copied exactly by you or anyone else. Those particular “dance steps” are organic to Kevin’s emotional energy and personality. Think of Gene Kelly, Fred Astaire, and Bob Fosse performing the exact same dance sequence for a movie. Even seen only in silhouette, so you couldn’t recognize their faces, there would be no question as to which man was dancing during each sequence. By the same token, everyone will do the heel exercise differently, depending on how they naturally express their own energy through their own physical and emotional centers-of-gravity. 

“Changing the World, One Dog at a Time”
Join Me on Facebook!
Follow Me on Twitter! Blog (archived)


*Breeds like Danes, Dobermans, Dalmatians, viszlas, boxers, greyhounds, who all have a similar chest conformation, are designed to run hard and fast for brief spurts, not to jog slowly for long stretches.
**Working on the heel the way Kevin does will use up more of a dog's energy in 5 - 10 minutes than a 1/2 hour jog will. If you add playing tug, fetch, and push-of-war, the dog's energy needs will be completely satisfied.