Four Ways to Walk a Dog, part 2
By Michael Lenehan part 1 | 2 | 3
WATCHING BERNIE BROWN teach a dog to heel would drive Bill and Dick Koehler up a wall. The Koehlers walk a dog on a slack leash and tempt it to err, so that it can learn from its mistakes; Brown holds the leash taut, so that the dog will have no choice but to walk in the correct position. The Koehlers forbid “bribing” a dog with food and “nagging” it with repeated commands; Brown teaches a puppy to watch him by fluttering his fingers at his belt buckle, sometimes holding a tidbit there, saying “Watch . . . watch . . . watch . . . watch.” The Koehlers think that dogs are smarter than most people give them credit for; Brown is one of the people who don’t give them much credit. “I treat a dog as a very simpleminded animal, like a mentally retarded child,” he says. “I go very slowly.” The Koehlers aim to teach a dog the “consequences of its own actions,” stressing the animal’s dignity and the handler’s authority; Brown tries to make training fun and exciting for the dog, expressing enthusiasm with a particularly gooey sort of baby talk: “Look at that baby! Look at that boy! Oh my goodness, oh my goodness gwacious, is that bootiful! That’s just scwumptious.” When he’s particularly pleased, he’ll hold a dog’s snout in his hands and kiss the dog full on the lips, loud and wet.
Finally, whereas the Koehlers are straightforward, perhaps to a fault, about the use of physical force in training, Brown’s attitude is a bit more complicated. Having learned his first techniques from traditional, military-style trainers—practitioners of what he calls the “jerk ’em, sock ’em, bang ’em method”—Brown now claims to have found a better way. His method, which he pointedly calls “the No-Force Method of Dog Training,” does not eschew force so much as seek to avoid it as long as possible. Brown admits that for some dogs forceful training is the only way to get results. But for others—the majority, he says—it does more harm than good. It can make a “soft” dog or a shy one nervous, withdrawn, even aggressive. Brown therefore envisions a ladder of escalating force, and he advocates taking each dog only as high as is necessary to get results. Where his own dogs are concerned, he’d rather replace the dog than climb all the way up, but of course he cannot ask the same from everyone who comes to him for help, so he operates by a double standard that he readily acknowledges. I once heard him say to a roomful of dog owners: “I can’t bang a dog around. I cannot live with myself if I have to beat up my dog. I can beat up your dog okay. I can handle that.”
At times the Koehlers and Brown appear to be in entirely different businesses, and in a way they are. Brown is one of the extremists of dog training—one of the “obedience people” (as distinct from “pet people,” in Brown’s parlance), who have turned training into a means of amusement, a form of human competition. Obedience competitions—or trials, as they are called—are regulated by the American Kennel Club and held at dog shows all over the country. In them dogs are required, among other things, to heel, jump over hurdles, retrieve wooden dumbbells, respond to silent hand signals, and pick from a group of objects the one that bears the handler’s scent. By accumulating points in these trials, handlers can work their dogs up through a progression of obedience titles, from C. D. (companion dog) to O.T.Ch. (obedience-trial champion).
Few people are better at this game than Bernie Brown. For three years running, 1979 through 1981, his male golden retriever Duster—properly Ch. and O.T.Ch. Meadow-pond Dust Commander—was the Ken-L-Ration Obedience Dog of the Year, meaning he had scored more obedience-trial points than any other dog on the circuit. Brown retired Duster in 1982, at the age of six. Until last summer, when he was surpassed by another golden retriever, Duster had scored more points than any dog in the history of AKC obedience competition. Such a feat requires not only a talented dog but also a driven handler. Brown, a former newspaper copy editor and magazine publicist, says that obedience competition is his ego trip. “You know, you walk into a room where there’s a thousand dogs, and you walk by and people say”—his voice drops to a whisper—“‘Look who’s here. That’s Bernie Brown!’ That’s my high in life. That’s what turns me on. Don’t think it’s not something to be able to walk into a room in Detroit or Carolina or Florida or California or wherever I go, and when I’m in the ring suddenly from all over the room people come just to stand by and watch—don’t think I’m not aware of that. Do you know I retired Duster a year before the dog should have been retired? Why? So nobody in the world, not one person alive, can say, ‘Oh, I saw that dog, he wasn’t so great.’ I retired him at his peak, because I never wanted anybody to say, ‘Well, Jesus, that’s Bernie Brown? What’s so great about him?’ I don’t want that. It would hurt my pride too much. I can’t handle that.”
Duster begat his own successor in the obedience ring, a male golden called Chip, short for Meadowpond Duster’s Chip Off the Block. (Meadowpond is the name of the kennel where the dogs were bred.) Chip was only sixteen months old when he competed in the Gaines 1984 United States Dog Obedience Classic, held in Elizabeth, New Jersey. He won in the novice division and earned the meet’s highest score, 199 out of a possible 200. But later, as Chip moved from novice to open competition, a spine problem surfaced, ending his career; the open level is where the jumping begins, and Chip was literally unable to make the leap. Brown found him a home with some pet people and began searching for another competition dog. He was still searching a year later. When I first met him he was high on Skor, a golden puppy that he had just begun to work with. A couple of months later, when we met again, Skor was in the doghouse and Brown had a new world-beater, Cajun. He said, “Regardless of how good I am—and I really think I am one of the best trainers in the United States—I am only as good as the dog working with me. Duster made me look fantastic. Skor makes me look like a piece of …. I think Cajun is going to make me look as good as Duster.” But it’s hard to judge a young puppy. By Brown’s own reckoning Cajun was the sixteenth dog he had tried as Duster’s replacement. His pride would not allow him to return to competition until he had found the perfect partner. “When I go into the ring, I want ’em to say, ‘That little . . . that sonofabitch has got another one.’ I want them to hate me.”
Success in the ring is Brown’s chief source of ego gratification; it also generates his chief source of income. He gives occasional group training classes, like the Koehlers’, and he makes house calls to deal with problem dogs, but most of his livelihood comes from private lessons and weekend seminars that he gears specifically to AKC obedience competitors. (He has also written a couple of training manuals, the first of which is dedicated to Duster.) I attended a seminar he gave in Bristol, Tennessee, for about fifty dogs and their owners, mostly women. (Obedience competition is overwhelmingly female.) He put on a good show. He’s a compact, peppery man of fifty-three, who could easily be mistaken for forty-three. He is affable, open, and hopelessly profane—if he fails to swear within five minutes of shaking your hand, you’re probably making him nervous. Working the floor of a YWCA gym with a wireless microphone clipped to his shirt, he instructed his audience the way he teaches his dogs: very slowly, repeating some sentences two or three times for emphasis. He larded his talk with “we”s and with teacherly phrases like “In other words” and “What do I mean by that?” Once in a while he brought Duster or Cajun to the center of the room to illustrate a point. (Skor had been left at home, in Hinsdale, Illinois.) Brown harangued the audience, swore at them, ridiculed their pitiful ignorance, and they laughed and ate it up.
“Write this down,” he told them. “As a dog trainer, you have one function only. One function only. What is that function? I’ll tell you. It is, you’re not to let your dog make a mistake. It’s that simple. What is your job as a dog trainer? Pure and simple, cut out the bull. . . Not let your dog make a mistake. Now, I am in the minority. Most trainers feel that we have to let the dog make a mistake so we can correct the dog. Because how does the dog know what’s wrong unless the dog is corrected? That’s the way ninety percent of the dog training is done today in this country. I think that’s wrong.”
An example of what Brown thinks is right is his heeling technique, which he calls binding. He uses an extra-short leash—he just happens to have them for sale; you can pay during the doughnut break—and grasps it just above the clip that attaches it to the choke chain. Holding the leash taut as he walks, he gives the dog virtually no room for error. He uses a different verbal command to correct each of the four mistakes the dog can make. If he feels the dog forging ahead, he commands “Get back”; if he feels the dog lagging behind, he says “Get up”; if the dog wants to swing wide to the left, “Get in”; if it crowds in toward his leg, “Get out.” Brown walks the dog in this fashion for minutes, days, months—until, as he likes to say, “the dog tells me it’s ready to go on to the next step.” The dog tells him through the taut leash, by not pulling in any of the four forbidden directions. The next step is to repeat the whole process, holding the hands a bit higher on the leash—one inch higher. Brown moves up the leash literally an inch at a time, until the dog can heel reliably on a slack leash and finally on no leash at all. Then it’s time to teach the right turn: back down to the bottom of the leash and up again, an inch at a time. This progression in tiny increments is characteristic of all of Brown’s training, even the advanced, off-leash work involving jumps and retrieves.
At the seminar Brown gave two main reasons for teaching the heel this way. The first stems from his working assumption on canine intelligence. The conventional way to teach heeling—used by Barbara Woodhouse, among many others—is to walk the dog on a slack leash, jerking in the appropriate direction to correct the faults of forging, lagging, and so on. The handler commands “Heel” on setting out and repeats the command with each correction—“Like this,” Brown mocked, jerking a poor imaginary dog on the end of a leash that he held in his hands: “If the dog goes wide—‘Heel!’ If the dog crowds in—‘Heel!’ I want the dog to go around—‘Heel!’ I want the dog to swing—‘Heel!’ Now look at that from the dog’s standpoint. Say we’re walking off-leash, where I can no longer jerk the dog, and I say to the dog ‘Heel!’ Should he swing, get up, get back, get in, get out—or fall on the floor in total frustration? I have given him six definitions for one word. So you say, ‘He can figure that out.’ Well, maybe your dog is smarter than my dog. See, my dogs are very dumb. So I give my dogs one word to mean one thing. ‘Get back’ means only one thing. ‘Get in’ means only one thing.”
The second advantage of this technique, Brown claims, is precision. “Jerking is a very unscientific way to get a dog in the heel position,” he said. “If I jerk the dog back, where do I jerk him back to? If I jerk him forward, how far is forward? Binding lets me bring the dog within a sixteenth of an inch of where I want him to be.”
IN THE STRANGE WORLD OF COMPETITIVE OBEDIENCE, inches count. To earn the title of companion dog, a dog must qualify in three separate novice obedience trials. In each trial it must score at least 170 points out of a possible 200, and it must receive at least half the points allotted to each of the trial’s several exercises. (For example, the off-leash heeling exercise is worth forty points; any dog that fails to score at least twenty on it does not qualify—a shameful fate known to obedience people as NQ.) Here is what the dog must do: heel on and off leash, staying in position when the handler speeds up, slows down, and turns; sit automatically whenever the handler stops; heel on leash as the handler walks in a figure eight; stand still and allow itself to be touched by the judge; come without delay when called and sit directly in front of the handler; move when commanded from this sitting position to the heel position, at the handler’s left knee, and sit there; sit and stay in a line of other dogs for one minute, and lie down with the other dogs for three minutes, while all the handlers stand on the opposite side of the ring.
To most pet people, a dog that could do all this would qualify as a fairly well trained dog. Most obedience people would consider a score of 170 reason to look for another hobby. Bernie Brown considers a score of 195 reason to look for another dog. The difference between the qualifying score of 170 and the perfect score of 200 lies in the speed, enthusiasm, and precision with which the dog completes the exercises. Points can be deducted if a dog lags a few inches while heeling, if it walks lackadaisically when commanded to come, or if it sits with its tail end a few degrees off center.
Brown thinks that competitive obedience is the most stressful endeavor a dog can endure—more stressful even than the responsibility of guiding blind people through busy city streets. When he first told me this, I scoffed, but later I came to understand what he meant. I suggested to him that the difference between a guide dog and an obedience dog might be something like the difference between an infantryman on patrol in wartime and a grunt performing drills in boot camp. The infantryman’s job is more dangerous and his responsibilities greater, but he enjoys a sort of freedom, and to an extent he can control what happens to him and his buddies. In comparison, the grunt in boot camp endures a particularly grating kind of stress. He has no freedom, no real responsibility, and very little appreciation for the utility of lockstep marching, snappy saluting, and precision bed-making. To the grunt, it all seems an arbitrary exercise of authority: If the object of marching is to get from one place to another, why am I being hollered at for walking out of step? Similarly, a dog might ask, if it could ask such things, If the object of sitting is to be still at this person’s side, why am I being corrected for placing my butt a few inches to the left of my nose?
Brown liked my analogy. When I went on to suggest that competitive obedience is perverse and unnatural, he said, “Probably, you’re not too wrong. I’ve said that dogs don’t think or reason; they communicate in animalistic terms that we don’t really understand. But I guess if dogs could talk, the dogs would say, ‘Hey, this is a bunch of …. A dog is not made to be a precision animal. God didn’t make him to be a machine, to be wound up like a robot.”
Brown, a pro among hobbyists, demands much more of his dogs than most obedience people demand—not only more precision but also more enthusiasm, or at least the appearance of it. In the ring he aims to present a picture of unalloyed joy, exuberance, communion between dog and master. His golden retrievers walk with an almost equine prance, bending their heads around his leg to look adoringly up at his face. It’s a sight to see, and Brown insists that it is a consequence not of training but of the dogs’ temperament. Once, for my benefit, he took Cajun for a spin around the ring; the dog was only fourteen weeks old at the time, but he already had that Bernie Brown look. Brown said, “The average dog walks along and says, ‘Aaah, let’s get this over with.’ But the natural temperament of this dog just says ‘I want to please you: you enjoy it, I enjoy it, I love ya, I’m happy, I want to do it, I’m having a ball.’
“Now,” Brown asked me, “did I train him to do that?” No, he had not yet begun to train Cajun formally. “The dog popped out of the womb like that,” Brown said. This is why he devotes so much energy to finding the right dog. He wants his dogs to look happy and enthusiastic under extraordinarily difficult circumstances, and he is convinced that dogs able to do so are born, not made. When he’s “campaigning” a dog, as he campaigned Duster for three years, he competes in sixty or seventy shows a year. “Load the dog into a car, drive five hundred miles one way to a dog show, compete on Thursday, Friday, Saturday, Sunday, take him back five hundred miles—week in, week out, week in, week out. You know something? In the history of the sport of obedience only three or four dogs have been able to handle that grind.” The grind is part of the reason that Brown tries to avoid forceful training. He says that if he showed ten times a year, he might be able to get away with using conventional methods. But, he says, “I know that a dog can’t stand that while competing seventy times a year. You can’t beat a dog up to make it work like that. Numerous competitive trainers have tried it, and they’ve all gone down in flames.”
IN HIS MANUALS AND SEMINARS BROWN STRESSES THAT whatever one’s objectives in owning a pet—competition, companionship, or anything else—selecting a puppy of the proper temperament is more important than any training method. Too few dog owners, he says, are aware that dogs even have temperaments, and their ignorance is the main reason that animal shelters are euthanatizing seven to ten million dogs each year in the United States. “You talk to a pet owner and say, ‘What about the temperament of your dog?’ and he won’t even know what you’re talking about. He’ll say, ‘I’ve bought a dog. My grandfather had a dog, my father had a dog, I had a dog, and now my son’s got a dog. It has two legs in front, two legs in back, and a tail. What’s temperament?”’ As he said this, Brown was warming up to a diatribe, getting into character. “Then the dog starts snapping, growling, barking, and it’s ‘What have I got here? What’s the matter? I better get something. I better get a book.’ The guy goes in the store and picks up Barbara Woodhouse’s book. ‘Hmmm, “There are no bad dogs, only problem trainers.” Jesus, I’m doing something wrong. Barbara Woodhouse, I’ve seen her on Mery Griffin. I’ve seen her on Johnny Carson. Obviously there’s nothing wrong with the dog—it says so here. There’s something wrong with me.’”
Brown believes that the world is crawling with bad dogs. The commerce in purebred dogs has increased dramatically since the end of the Second World War. In 1930 and 1940 the American Kennel Club recorded about 48,000 and 83,000 new registrations respectively. By 1950 the number had shot up to about 252,000, and in 1970 it exceeded a million. (The number of purebreds registered by the AKC is of course only a small percentage of the country’s total dog population. According to the Pet Food Institute, the United States is home to about 50 million pet dogs in all.) Brown believes that much of the rising demand for dogs is being met by less-than-conscientious breeders and pet-shop operators; the result, he thinks, is a serious decline in the temperamental quality of our pet-dog population. “In 1940 I had my first dog, a cocker spaniel. The cocker spaniel was the number-one breed in the country. Everyone wanted a cocker spaniel because of a dog named My Own Brucie. He was a black cocker, and he was on all the cereal boxes and magazine ads. . . . The breeders went insane. They bred mother to son, father to daughter, and on and on, and suddenly the cocker spaniel turned into a vicious, snarling, mean, aggressive little bastard. Today the cocker is one of the most miserable-temperament dogs in the world.”
Almost all of the dog people I know agree with Brown that sloppy breeding—and particularly the overbreeding that tends to occur when a breed becomes very popular, as a result of a movie or a TV show, for example—has seriously harmed some breeds, both physically and temperamentally. Most would agree that a prospective pet owner can avoid many common pitfalls by educating himself, getting good advice, and carefully selecting a dog rather than simply obtaining one. This does not mean, however, that you have cause to panic if you’ve recently come home with a hastily acquired cocker spaniel. Breed stereotypes are like any others: valid though they may be as generalities, they do not apply reliably to individual dogs. Moreover, Brown’s perspective differs greatly from that of the ordinary pet owner. As a competitor he demands much from his dogs, as a trainer he sees a disproportionate number of incorrigible cases, and as a character he tends to express himself enthusiastically. The cocker is not the only dog whose temperament qualifies as miserable in his view.
“The same can be said of the Irish setter,” Brown says. “They bred the brains out of the Irish setter. They bred the brains out of the German shepherd, the Doberman. They’re breeding the brains out of the rottweiler, and they bred the brains out of the collie. The dog your grandfather had is not the same as the dog your son has. The collie had a big, wide, pie-shaped head, and he was massive, a hundred pounds. Now the collie’s head is about that narrow”—Brown was holding his fingers a few inches apart. “The old Irish setter had a massive, gorgeous head. Today an Irish setter is an insane, off-the-wall maniac. They bred the brains out of him. That’s where dogs are going today.”
DANIEL TORTORA IS THE ONLY DOG TRAINER I MET who was able to say, with a perfectly straight face, “Come on into my office.” His office is on a wooded estate, about three acres, in Upper Saddle River, New Jersey, an expensive suburb of New York City. His house is just outside the office door. His pool is out back, and his fire-engine-red Mercedes is usually parked up at the top of the hill. As we talked in one corner of the office, Tortora’s assistant, Andrea Meisse, was working in another corner, entering data into a computer system that keeps track of Tortora’s cases. Shelves in the room are packed with video cassettes documenting some of his more impressive cures. Behind his large desk is a small library of academic and reference books—Aversive Conditioning and Learning, Physiological Correlates of Emotion, Structure and Function in the Nervous Systems of Invertebrates, and dozens more. Arrayed on one wall are a collection of diplomas, including a Ph.D. in psychology from Michigan State, 1973. A sign down at the roadside identifies the place as Dr. Tortora’s Pine Hill Kennels. A few miles away in Queens, where Tortora grew up, it would be called a nice setup.
In Tortora’s training method—which is named RemBehCon, for remote behavior control—a dog is taught to heel in eighteen discrete steps (nineteen if you count heeling backward), enumerated in a computerized document called the RemBehCon Training Thesaurus. The first step, command level 1, goes by the description INDUCED HEEL. Definition: dog can be induced to follow handler, walk alongside handler. The inducement might be a tidbit, a hand slapping playfully on a thigh, or merely the dog’s affection for the handler. Command level 18 is SUCTION: find the heel position from progressively increasing distances from handler with an attraction close to the dog + FULL HEEL. (FULL HEEL is command level 9, a drill-review of levels 1-8.) “This is what most people have, if they’re lucky,” Tortora told me, pointing to the top of the list. “This is what most people want,” he said, sweeping his hand down toward the bottom. He laughed generously, as he often does; the ironies of life and of dog training seem to amuse him greatly. “Customers come to see me and they say, ‘I want something very simple. All I want the dog to do is to listen to me when I tell it something.’ What they mean is ‘All I want the dog to do is come when called when he’s chasing a rabbit, or sit calmly when he’s getting prepared to jump excitedly on a guest.’ All they want the dog to do is be controlled in the most demanding situations.”
As Tortora defines it, dog training is the act of getting from what the customer has to what the customer wants—from the top of the list to the bottom. His Thesaurus, which he says is one of the keystones of his method, lays out for each of eighteen different commands a progression of tasks arranged in order of increasing difficulty. A dog is not asked to heel at level 6 (circles and figure eights) until it has mastered level 5 (straight lines at varying paces). In this respect Tortora is similar to Bernie Brown. “A correction trainer will be around here,” Tortora said, waving his finger over the middle of his list, “and then the dog will be exposed to an attraction, and it will be punished, or corrected, for going toward the attraction. In my way of training that’s sort of like teaching a child arithmetic, giving it a test in calculus, and then punishing it for failing.” Bernie Brown’s cardinal rule of training—never let the dog make a mistake—is in Tortora’s argot “errorless learning,” a concept that he says is well accepted among behavioral psychologists. And Tortora’s leash technique—“leash pressure,” he calls it, to distinguish it from the traditional jerking—is somewhat similar to Brown’s binding. But Tortora parts company with Brown, and many other trainers, when it comes time to move from on- to off-leash control, probably the most difficult step in all of dog training. This is the point where Bill Koehler resorts to the throw chain and the fishing line. Tortora has used Koehler’s techniques and admits that they work. “But you can’t imagine the cursing,” he says. “You know, if you’ve got a fish on the end of a fishing line, it gets tangled up. You can just imagine the potential problems if you’ve got a dog on the end of the line.” Tortora, as one might expect from a quick look around his office, prefers a higher-tech approach. What puts the Rem in RemBehCon is another keystone of his method—a Tri-Tronics model Al-90 remote-control trainer, or what is commonly called a shock collar.
Tortora’s kennel is a place where nitty-gritty dog training meets highfalutin experimental psychology. That’s a rare intersection. Here Tortora is advising the readership of Gun Dog, “The Magazine of Upland Bird and Waterfowl Dogs,” on the qualities that make a desirable retrieving prospect:
I want an easily developed natural fetch. I want the pups to chase after a bumper when it’s thrown and have a desire to carry it around. I want pups that will naturally choose water over land.... If you find a pup with a good set of qualities, you can get a shallow water retrieve in his first session with water. I got all of this and more in my pups, so I’m pleased. So now what?
Here he is addressing his colleagues in the Journal of Experimental Psychology: General, 1983, Vol. 112, No. 2:
The data seem to suggest that safety training may create in dogs a sense of control over environmental stressors. By teaching the dogs a behaviorally balanced battery of prosocial “coping” responses, they may be developing the canine counterpart of “self-efficacy” or “courage.” This cognitive modification . . . may be of prime importance to the success and stability of the results.
LIKE MOST OF THE DOG TRAINERS I’ve encountered, Tortora trained his first dog in his early teens, largely by the seat of his pants. “He was a combination German shepherd and yellow Lab, a really big sucker. I have no idea how I trained him. I just figured it out. I never read a book on it; I think I read a pamphlet, but I don’t think I heeded it. It just made sense to me to do it this way: give a command just before the dog was about to do what I wanted him to do anyhow.” Later, as a psychology student, Tortora would learn a name for this training technique—overlaying. He used it to teach his dog to defecate and urinate on command. “I waited until he was about to have to go anyhow; then I’d run him outside and give the command for defecating, which was ‘Hurry up.”’ Tortora told me he went to camp one summer and left the dog with his father, neglecting to give complete instructions. His father walked the dog ten times in two days, to no avail, and finally called the camp in a panic. Tortora told him the secret words, and the crisis passed.
At Michigan State, Tortora worked not with dogs but mostly with rats, and mostly in the area of what is called escape-avoidance learning. He gave me a brief lecture on it from behind his desk, stopping occasionally to fish for a journal reference or to pull a book from the shelf behind him. (He teaches a few psychology courses at Jersey City State College.) “In a standard, basic escape-avoidance study—the first one was done in 1948, by a guy named Neal Miller, but his experiment was a little more complicated than this—you’d put a rat in a box. On one side of the box there are shock grids on the floor, and on the other side there are none. The two sides are divided by a small hurdle. A light comes on, and a second later electrical stimulation is sent through the floor. The rat hops around—no attempt is made to make this a nice experience for the rat—and then jumps over to the other side.” With repeated trials, Tortora. said, the rat learns to jump to the other side of the box as soon as the light comes on, before the electrical stimulation begins. “So the rat goes from escaping the stimulation—that is, jumping when it is on, when he’s sort of having a hotfoot—to avoiding it. Now, once you teach a rat a really clear and unambiguous avoidance response, it sticks; it’s very hard to extinguish that habit.” Many behaviors learned by lab animals fade away when the reinforcement used to elicit them, whether positive or negative, is removed. If pecking a key no longer yields a morsel of grain, the pigeon soon loses the inclination to peck. If turning the wrong way in a maze no longer causes its foot to be shocked, the rat soon ceases to go the right way. In short, when the rules of the game change, the animal’s behavior changes as well. But in avoidance learning the animal does not know that the rules have changed. “If you turn off the electricity,” Tortora said, “so the light comes on but there’s no shock following it, the rat still jumps. It makes sense, because the rat’s jumping before it can find out that there’s no electricity anymore.”
In Understanding Electronic Dog-Training, a book written for and published by the manufacturers of the Tri-Tronics training collars, Tortora explained the principle of avoidance learning in dog-training terms:
Research has proved that a habit learned through avoidance training is the most permanent of all habits. To understand why, look at the situation from the dog’s perspective. Let’s say you trained your dog to sit using avoidance training. This means that you have convinced your dog to expect that sitting on command prevents discomfort. Thus, every time he sits on command, he feels that he has successfully avoided the discomfort, regardless of whether he really did or not. Everytime he sits on command, he reinforces his belief that the rules are still in effect since he didn’t feel electrical stimulation.
There’s an old joke that conveys the powerful effects of avoidance training. It goes like this: A man walks into a psychiatrist’s office and sits on his hands. When asked by the psychiatrist why he is sitting in such a peculiar position, he says, “I always sit on my hands, it keeps the lions away.” The psychiatrist replies, “There are no lions in Manhattan.” To which the man responds, “See! It works.”
After the dog has been trained to sit on command with avoidance training, if he could be asked, “Why are you sitting so nicely when commanded, you haven’t felt electrical stimulation in years?” the dog would reply, “See! It works.”
Tortora began working with dogs again after he completed his doctorate, in 1973. Within about a year he had hung out his shingle as an animal psychologist in Spring Valley, New York, not far from his present home. But he did not immediately understand how relevant his academic training was to the problems that pet owners were bringing him. When he first encountered the electronic dog-training collar, he didn’t even see the connection between it and the shock grids that he had worked with in grad school.
“I was taking dogs in, and I was running up against problems like, Here’s a dog doing something wrong when I’m not there, and I need to be able to stop it. I think stealing food was the first case like that. What I needed was a device to deliver an aversive stimulus when I wasn’t around. But I didn’t even know that radio-control collars existed.” At the time, a few shock collars were on the market; in principle (though not in electrical output) they were basically remote-control cattle prods, popular especially with the trainers of tracking and retrieving dogs, who need a way to correct their dogs at great distances. Before such devices were developed, some hunters used the fishing-line method that Bill Koehler advocates. Others used such imaginative techniques as “dusting,” which is essentially shooting a load of buckshot at the ground near the dog, so that the earth appears to explode forbiddingly in response to the dog’s mistake. Tortora sought something a little more practical. “I made a few calls,” he said, “and bought my first radio-control collar from a trainer who said, ‘Oh yeah, I got one of those.’ He dusted it off—I think he was grateful to find somebody to take it off his hands.” Tortora soon learned why. “When I first used it, I used it the same way everyone else was using it—to punish the dog. And I quickly saw the problems. The intensity was too high, and the results were not predictable. If a dog was jumping up on the kitchen counter, I could fix that problem using the collar, but then the dog wouldn’t even go into the kitchen anymore. Or if the phone rang just before I pressed the button, the dog would be afraid of the phone. These were unpredictable results you wouldn’t get when you had an animal in a Skinner box, where there was only one thing it could do.”
Tortora said he floundered with the collar for about a month before realizing, suddenly, that he had been seduced by the ethos of traditional dog training and had abandoned the principles of behavioral psychology. “I remember sitting in my old office, in Spring Valley, and doing one of these,” he said, smacking his hand against his forehead. “I said, ‘What, did I forget what I’ve been doing for the last seven years?’ Then I started thinking about how to use the collar in the escape-avoidance paradigm, and everything grew out of that.”
Tortora began talking with the collar’s manufacturers, who were in the process of transforming their product from a dirty little secret into a rather sophisticated behavior-modification tool. The original Tri-Tronics collar was fairly simple: pushing a button on a hand-held, battery-powered transmitter sent a radio signal to a box on the dog’s collar; two pronglike contacts protruding from the box conducted an electrical charge to the skin of the dog’s neck. The charge was supposed to hurt, just as a jerk on a leash is supposed to hurt; the purpose was punishment, and no one pretended otherwise. Present models are different. Now the user can set the electrical stimulation at one of five levels, the highest of which corresponds to the single level available previously. Tortora, who helped the company determine which levels would work best, believes that the change allows the collar to be used as a motivation rather than a punishment. At a low level, he says, the electricity may stimulate the dog’s neck muscles, causing them to tense, without activating the pain nerves, whose response threshhold is higher. Thus the stimulation may be felt not as pain but as a “pressure to perform,” similar perhaps to the pressure that a dog feels when a choke collar binds it in a heeling position. Tortora has some reason to believe that when a dog is naturally motivated to perform—for example, when a retriever is fetching, indulging the instincts bred into it over centuries—the stimulation may even be felt as a form of arousal. Both Tortora and John Maurice, a trainer who was working for him when I visited his kennels, told me that a dog’s perception of the electrical stimulation can depend on the way the handler presents it. Tortora demonstrated this to me with one of his own dogs, Levi, a German short-haired pointer. He gave her a jolt while she was running full-tilt after an airborne Frisbee; she ran right through the stimulation, with no reaction that I could discern. A few minutes later she made a mistake in another exercise. Tortora yelled “No!” and gave her another jolt of equal intensity. She yelped in pain.
(I have felt the third strongest of the five available shocks on my finger, which is no doubt more sensitive than a dog’s neck. The sensation was certainly uncomfortable, but I’m not sure I can call it pain. I have seen the collar used on several different dogs by several different handlers, at Tortora’s kennels and elsewhere. Usually the dogs flinch or tense visibly in reaction to the stimulus, but they do not act as though they have been struck or punished.)
Two of the three remote-control trainers now available from Tri-Tronics are designed explicitly to serve as escape-avoidance teaching devices. In these models each burst of electrical stimulation is preceded by a conditioning tone audible to the dog—a brief buzz that predicts the onset of stimulation, just as the light does for the rat in the shock box. Like the rat, the dog can learn to avoid the shock by responding quickly to the buzz. A button on the transmitter allows the handler to send the buzz only, without any accompanying electrical stimulation. In addition, the top-of-the-line Tri-Tronics trainer also produces a second tone, a beep that follows the electrical stimulation. This feature, added at Tortora’s instigation, is based on “relaxation theory,” a concept developed at Michigan State by Tortora’s major professor, M. Ray Denny. The beep serves as a “safety tone,” an indication that the shock is over. It too can be decoupled from the electrical stimulation, and can therefore be used to signal the dog that it has performed well and has avoided the aversive stimulus. “From half a block away,” Tortora says, “I can say ‘Good dog’ by pushing a button.” Eventually, if the training goes according to plan, the electricity drops out of the system and the dog can be trained entirely through the use of the conditioning tones.
In “Safety Training: The Elimination of Avoidance-Motivated Aggression in Dogs,” a study he published in the Journal of Experimental Psychology: General, Tortora describes using the remote-control collar to train thirty-six problem pets—biters, snappers, and growlers referred to him by veterinarians. These were what he calls “avoidance-motivated aggressors”—dogs that in his judgment had learned to use aggressive behaviors as means of avoiding stressful or unpleasant experiences. (Rambo, the kitchen terrorist invented for the beginning of this article, would probably qualify as an avoidance-motivated aggressor.) Tortora trained these dogs to respond to the standard obedience commands—come, heel, sit, stay, and so forth—according to his usual procedures. That is, he introduced the electronic collar only after each dog had learned the commands through more conventional training methods, and he played fetch or tug-of-war with the dog as a part of the training regime, because he believes that such play—the only chance most dogs ever get to exercise predatory instincts, which they no longer really need—is the most powerful reward a trainer has at his disposal. (It is also, he points out, more convenient than a pocketful of smelly liver treats.) In the end, Tortora claimed, every one of the thirty-six dogs was cured completely. Having learned aggression as an avoidance response, he explained, they were able to discard it, or suppress it, after learning a new set of avoidance responses in training. By imposing stress on the dogs, in the form of electrical stimulation, and teaching them that they could avoid it by obeying commands, training gave them alternative ways of coping, and ultimately a sense of control over their environment. Even their carriage—“the manner of holding and moving the head or body”—improved. In a word, the dogs became confident.
THE TRI-TRONICS COMPANY, which retains Tortora as a consultant, has taken pains to inform dog people that its products seem to be most effective when used as behaviorist tools rather than as instruments of punishment. But prejudice against the shock collar dies hard. I have encountered several dog trainers who dismiss it as vile but who apparently don’t understand how the manufacturer recommends using it. I have heard the editor of a dog magazine announce that he’d never publish an article encouraging its use. Tortora believes that the prejudice is rooted in the mind-set of traditional dominance-submission dog training: trainers who know only punishment and physical dominance naturally assume that these are the purposes for which the collar must be used.
Tortora thinks that dominance does enter into training, but as a consequence, not as a starting point—something a trainer earns rather than something he seizes. A well-trained dog perceives its teacher as one who “gives good advice,” Tortora says. “When I say something and the dog does what I say, or if the dog makes a certain movement after I make a certain sound, the dog eliminates pressure and gains reinforcement. If you consistently give the dog good advice about how to deal with its environment, then of course you’re going to be the leader. If you think about the head of a pack, a dog or a wolf may bite or attack the subordinate members if they get out of line, but I don’t think that’s the key to holding dominance. I think the key is successfully controlling the whole pack’s movements. If you ever saw a pack hunting, you know that the way the leader moves controls the rest of the pack. The result of that of course is the chase, or the successful bringing down of the prey. So it’s movements—not obedience, not right and wrong, and not necessarily dominance. You never seek that, you get it. It’s a by-product.
“If you look very carefully at the successful correction trainer, you’ll see that he’s using the terminology of correction training but he’s actually doing escape-avoidance learning. I’ve watched, I’ve had the opportunity of being sent around the country to watch very successful trainers. They use the same words as other trainers, but they’re a little out of kilter. They’ve learned to do it a little different because they say it gives them a ‘better result.’ They have discovered escape-avoidance learning.” So have horse trainers and elephant trainers, Tortora says. When a rider wants a horse to turn, he applies pressure to the head or mouth, and the horse turns to escape the pressure. A mahout trains an elephant the same way, using a clawlike device to pull the animal’s ear. “Horses have a dominance hierarchy,” Tortora says. “Elephants do too. But I don’t think the mahout seeks to dominate the elephant. Ultimately he does dominate, but—you know, with a dog the suggestion is you throw it over on its back. What do you do with an elephant?”