Promoting Positive Training Methods
Anything you can teach a dog to do including barking and biting can be trained with positive reinforcement methods.
Every so often, at a training demonstration or event promoting positive training methods, a skeptical spectator will ask me whether positive training methods can be used for preparing dogs for all types of careers. I know where they are usually going with this question. Their real question is, “I know you can teach dogs to do cute little tricks with treats and stuff, but what about when you want a reliable dog, like an obedience competitor, a protection dog, or a police dog?”
Their assumption is that in order to teach a dog to respond without fail, to sharply execute the handler’s every command, you will have to use force- and fear-based methods at some point in the dog’s education. When circumstances dictate that the dog has to do what the handler wants, you’re going to have to train that dog to be afraid of what might happen if he doesn’t do it, right?
Wrong, wrong, wrong.
Why “protection” dogs are no different
Police and military dog training are just two of several applications of protection dog training. Others more likely to be encountered by the average dog owner include personal protection, plant security (guard dog), and protection dog sports such as Schutzhund and Ringsport.
Some dogs are taught to simply detect an intruder and sound an alarm (barking). These dogs are useful in situations where biting, or even the threat of biting, could not be permitted. Some dogs are trained to give alarm and offer some physical threat to the intruder. A threat dog must be willing and able to stand his ground after barking at an intruder. Other dogs are trained as true “man-stoppers,” willing and able to win a fight with a human, if necessary.
Dogs trained for protection sports must also demonstrate a high degree of competence in basic obedience training. All of these disciplines traditionally use a significant amount of compulsion and punishment in their training programs.
It’s taken a while, but the fun and effective positive training methods that have produced untold thousands of happy, confident, and obedient pet dogs are finally finding their way into the hands of the trainers who produce police and protection dogs. Increasingly, these enlightened trainers are realizing that positive training not only produces a dog who is just as reliable as a dog trained with old-fashioned force-based methods, but also builds a strong and deep relationship between dog and handler.
“Old school” dog trainer learns new tricks
Interestingly, some of the most vocal advocates of the positive methods in police and protection work are veterans of the “old school,” where dogs are physically forced to comply with the commands given them. That is definitely the case with Seattle, Washington, trainer Steve White, who used to “string up” dogs with the worst of them. But not anymore. In a law-enforcement world populated by tough men and women who use harsh training methods on even tougher dogs, this energetic, fast-talking police officer is a beacon of enlightenment.
Like many positive trainers of today, White got his first dog training experience by attending an obedience class with his own dog, in 1974. The perceptive instructor of that class recognized White’s natural talent with dogs and invited him to become an assistant. White did so, then subsequently joined the military and attended patrol dog handler’s school at Lackland Air Force Base in Texas.
It was in the military that White perfected his punishment-based training methods. From the beginning, he was taught to never back down from a dog, under any circumstances. The first dog assigned to him tested the limits of this type of program. “Astro” was distrustful and aggressive with his handler from their first encounter.
The pair had an armed truce until week four of the training program, when the “Down” exercise was introduced. The military’s method of teaching the down was to tighten the choke chain and drag the dog to the ground. Astro resisted, and the battle was on.
“There was a horrible melee of snarling, teeth, saliva everywhere . . . I felt the choke chain slipping in my hands and knew I was losing ground. So, as per instructions, I strung him up to near-unconsciousness, yelling ‘No! No! No!’ all the while, then tried to down him again when I felt him weakening from lack of oxygen. Meanwhile my Air Force instructors were screaming expletives at me that impugned my manhood, my intelligence, and my family tree.
“I got the same results with my second attempt. I strung him up again and spun around, doing what is known as a ‘Touch and Go.’ I tried another down. This time, amid the saliva, teeth, and snarls, I felt a distinct ‘Oh, that hurts!’ pain sensation. When I went to string Astro up the third time, blood flew everywhere. My instructors quit yelling. Once Astro quit fighting I quit stringing. I got a reprieve from the training program while my wound healed.”
When White got back to the training program two weeks later, he had a new dog, new instructors, and he completed the course without further incident as a Distinguished Honor Graduate. Meanwhile, Astro was finally certified – after biting both subsequent handlers – and shipped off to Morocco, where his fate is unknown.
Seeking other methods
White was highly successful by the standards of his military unit, but he was dissatisfied with his military dog training experiences. While both the Koehler method he had learned from his first instructor and the military methods were effective for some dogs, White realized that they certainly didn’t work well for all of them, as Astro had demonstrated.
Post-military duty, White went to work for the Kitsap County (Washington) Sheriff’s Department. He was assigned another tough dog, and started dabbling with food and toys as part of his training program. He soon got a reputation for having dogs who could do tricks. However, he says that when push came to shove, and when he had to get a dog to do something, he fell back on the use of force.
White continued in that way for some time, using non-compulsive methods for the stuff that didn’t “matter,” and having a lot of fun with it, and using more serious methods for more serious work. Then he began to realize that there was some bleed-over into the work aspects of his training. His trick-trained dogs seemed to have a better attitude about their work.
White continued to search for a method that would work for all dogs, amassing a huge library of training and behavior books in the process. One day in the early 1990s, while working for the Seattle Police Department as a part-time narcotics-dog handler, part-time trainer, and teaching seminars around the country, he found a small tan paperback book that changed his life. While searching for reading material for a flight home from a seminar, he found Karen Pryor’s landmark book, Don’t Shoot the Dog, in the self-help section of a bookstore. When he got on the plane he settled into his seat, opened the book, and was immediately and totally absorbed. The book deeply resonated with him.
“Karen Pryor showed me that while I thought I was looking for a perfect method,” says White, “what I really needed – what every good trainer needs – is an understanding of principles – simple principles that apply to all learning.”
Pryor’s book included three lists of training principles: “The 10 Laws of Shaping,” “The 4 Conditions of Stimulus Control,” and “The 8 Methods of Getting Rid of a Behavior.” White had these 22 sentences printed on a card that he carries (and gives to students) to serve as his guide to handling any training challenge.
Spreading the positive word
Since his Don’t Shoot The Dog epiphany, White has moved and started his own company, Professional Training Services in Seattle. He now spends much of his time providing handler and animal training and consultation services to government, industry and private parties around the country.
White works with a lot of law enforcement agencies, and has credibility where other positive trainers might not, because of his background in and understanding of police work.
As can be expected, he still meets with resistance from force-based training traditionalists, although he reports that it’s getting better. The fact that White can get great results training police dogs with positive methods plays a key role in his increased acceptance. An almost universal problem with trained police dogs is their frenzied barking behavior in the police car.
“I recently returned from the United States Police Canine National Seminar in Boulder, Colorado,” says White, “where we did a lot of work on problem solving with the barking-in-car behavior. Traditionally, police officers have tried to eliminate this behavior through punishment – pulley systems to drag the dog to the floor of the car when he barks, electric shock collars . . . Often this results in a more frenzied barking response from the dog that escalates to spinning and yelping.
“We worked with shaping the absence of the behavior in the dogs – using a ‘Good!’ and a treat when the dog was quiet. If the dog’s threshold (the distance at which the barking was triggered) was 150 feet we started at 160 feet and gradually worked closer and closer. We made progress with every single dog there using this method. By the end of the week, I had given away all my shaping cards and the 500-plus clickers I had brought. Some handlers commented that we had accomplished more in a half-hour of positive reinforcement than they had in years of punishment.”
The ultimate challenge
While most dog owners spend a lot of energy convincing their dogs not to bite, police, Schutzhund, and other protection dog work encourages the dog to bite; it’s a requirement of the job description. As it turns out, the hardest part of this training process is not training the dog to bite – most dogs enjoy the bite work! The most difficult and critically important part of bite work is convincing the dog to let go on command. Because a protection or police dog finds the bite work highly self-reinforcing, the release (or “out,” as it is usually called in the trade) can be a difficult behavior to obtain on cue.
Traditionally, the “out” has been taught using force. William Koehler, a highly respected trainer in his day, describes these methods in his book, The Koehler Method Of Guard Dog Training, published in 1962.
Koehler’s first approach to the “out” involves the use of a throw-chain and sharp jerks on a choke collar. If these fail, he advocates using a wooden dowel that has been slipped inside a rubber hose with which to give a “good clean chop midway across the top of the dog’s muzzle should he fail to ‘out’ immediately on command.” If that fails, he gives happy approval to the use of the electric shock collar. Koehler’s methods are still followed by many compulsion trainers doing protection dog work today. Even many trainers who concede that positive training can work in some areas often insist that you cannot train a positive release.
White disagrees. He teaches students how to train dogs to “self-out,” a positive method that puts emphasis (like all positive methods) on rewarding the dog when it lets go. White says the self-out has an excellent history of success for people who use it correctly, but adds that he doesn’t often see it used correctly.
“The self-out has gotten a bad rap because people have tried to use it who don’t understand the importance of positively reinforcing the release and working with the dog and his level,” he says.
In a number of past WDJ articles about positive training methods, we have described how any behavior can be “put on cue” if a trainer is observant, patient, and swiftly offers the dog an irresistible reward at the very moment the dog performs the desired behavior. Even though teaching a dramatic behavior like biting (and letting go) seems like it would call for more dramatic methods, the self-out can be taught in the exact same way as “garden variety” tricks.
First, White establishes that the dog is thoroughly prepared in the preliminary steps of training. He especially wants to see that the dog is equipped (so to speak) with what he calls a “default behavior,” a requested activity that the dog is highly likely to comply with. His favorite is “Down.”
Using food and play with the dog’s favorite toy, White spends as much time as it takes to ensure that the dog responds with the Down behavior with a high ‘probability of occurrence’ – meaning that the dog is very likely to execute the behavior under even difficult circumstances, or when he doesn’t know what else will pay off.
Then the dog is sent to engage the training assistant’s protective sleeve. Once the dog bites the sleeve, White simply waits. He gives no commands and offers no encouragement; neither does the assistant resist the bite. They just wait for the dog to let go, or to begin to let go. The dog is rewarded instantly if she does happen to let go, but this often takes a while on the first few tries, says White.
“The first time I did this with my dog I waited 18 minutes with her hanging on my arm, until I could see her start to waver,” he says. “At that moment I took out her favorite toy – a piece of fire hose – and gave her the ‘down’ cue. High probability of occurrence – at that point in time she was willing to down, and since she couldn’t comply with down and keep the sleeve in her mouth at the same time, she let go.”
White immediately cues the dog that she has done the right thing by making a “Click!” sound with his mouth (it’s hard to do all this and hold a clicker, too!), and giving the dog her toy. He also adds an additional positive reinforcement by giving her a cue and allowing her to take a new bite.
White follows the same procedure with each successive bite, reinforcing each self-out with a Click!, her toy, and another opportunity to bite. Eventually, he adds a verbal cue for “Out” to the mix, as the dog makes the connection between the new cue and the desired behavior. As long as the reward for the Out behavior is appropriately timed and just as (if not more than) enjoyable as the biting behavior, the dog will quickly and reliably learn the behavior – without stress and without violence.
A word on punishment
Most of the people who train dogs for police work – even the ones who use positive methods, like White – feel that there is a time and a place for the use of punishment.
“Cops are results-oriented,” White explains. “If there’s a problem they want to fix it right then and there. They are also professional fault-finders, who spend their working lives trying to catch people doing something bad. We often have a hard time getting them used to the idea of catching the dog doing something good. To make matters worse, punishment is also reinforcing to the punisher – we get an instant response and that makes us feel good – which is why punishment-based training sticks around even when, like shocking the barking dogs in police cars, it frequently doesn’t give us the long-term result that we want.
“By far the best approach to training is the positive one,” White continues. “Still, on the street, we must have a way to stop the dog who makes a wrong decision if he doesn’t respond to our command. Otherwise people, or the dog, can get hurt.”
For these reasons, White uses what he calls a “conditioned punisher,” that he varies depending on the individual dog. He might use a throw chain for one, a bonker (rolled up towel), a verbal reprimand, or leash correction, but what they all have in common is that he uses the punishment very sparingly. “If you use a punisher too often you create a punishment ‘callus.’ If the dog becomes inured to the punisher and you have to use it harder and harder for it to be effective, it’s a clear indication that you are doing far too much punishing and not enough reinforcing,” he says.
Changing training traditions for the better
White reminds us that a dog’s basic philosophy of life is “What’s in it for me, now?” The rest of us, including cops, aren’t all that different. If we can show law enforcement that the “what’s in it for me” in positive police dog training is a better-trained, more effective, better-behaved police dog – a more reliable and valuable partner for them in the field – police dog trainers will flock to the positive training camp.
Thanks to Steve White and the growing legions of positive dog trainers, as more officers realize the value of having a positive relationship with their K9 partners, the lives of police dogs across the country will be greatly enhanced, until we see “positive police and protection dog training” become more widely accepted and even less of an oxymoron than it is today.
Pat Miller is a freelance author and a professional dog trainer in Chattanooga, Tennessee. For information about her training classes, see "Resources."