Discipline: Drawing the Line
Peggy Moran


       Are you training your dog, or planning to? Perhaps you have enrolled in a class, or are being helped by a private instructor. Or you may be attempting to educate your dog on your own, with the help of books, video tapes, or training memories from your past. Most people with pet dogs have at least a few training goals; some no more lofty than puppy "kindergarten"--housetraining, basic obedience, and simple social manners; while others aspire to canine "graduate school"--advanced obedience, competition, or performance training. Rather than just hoping their canine companions "figure things out" on their own, most dog owners expect to offer at least some educational guidance to their pets. Goals vary from one person to the next, but generally fall under one of three categories: housetraining, behavioral modification, and performance (or obedience) training. Done correctly, training increases a dog’s confidence, social skills, self control, controllability, overall quality of life, and even life expectancy. Trained dogs can go more places, enjoy more privileges, and be safer than untrained or improperly trained dogs. Because very few pet owners are totally confident they possess the skills needed to most effectively train their dogs, many seek outside assistance.

       When searching for a dog training program, pet owners often investigate various training schools and instructors in order to determine which will be most suitable for themselves and their pets. If, during an initial visit, owners were to observe what they perceived as excessive use of the "wrong" training tools and methods, they would be unlikely to enroll in that particular program, continuing their search for one which was suitable.

       So, what exactly are the right tools, and what makes one training system more suitable than another? Based on the wide and contrasting variety of methods and opinions available for owners to choose from, one would conclude it must be "whatever works." But when it comes to training, the end doesn’t always justify the means, especially if "the means" hurt, and fail to truly educate the student. When it comes to discipline, are there lines one shouldn’t cross, in the name of accomplishing training goals?

       Discipline is a touchy subject in our society, and very strongly biased opinions can be found coming from two corners. The "negative training" school of thought focuses upon methods for teaching dogs (children, etc.) what they should not do, and uses principals of punishment and avoidance to discourage inappropriate behaviors. The "positive reinforcement" school of thought prioritizes teaching students what they should do, to earn positive results. Positive training systems emphasize the point that an individual engaged in a desirable behavior cannot be simultaneously behaving inappropriately.

       What a dog owner will consider the "right" or "wrong" way to discipline varies from one individual to the next, based upon their personal opinions, experiences and education. Many pet owners bring their own childhood disciplinary experiences and observations into adulthood, and then employ these approaches with their own pets without questioning their validity. "This is how my parents did it, and our dog turned out OK" is a common explanation for a particular approach being recycled from one generation to the next. Many pet owners do seek an educated "better way," leaving family traditions behind. They train their dogs using methods and tools suggested by someone they respect as more expert in opinion than themselves--a breeder, a veterinarian, a friend, a dog book, a trainer--adhering to them diligently even when they do not seem to be working. The most difficult task for the pet owner who hasn’t made learning theory a subject of research and study is figuring out who really does understand the best ways to train a dog, and determining which advice to take and which to leave.

       Everything a trainer needs to know about the science of training and how dogs really do and do not learn has been available in published books and papers for quite some time now, but the science sometimes seems to get lost behind the fiction. The "good" fictionalized dog thinks and feels like a human, but is kinder, nicer, more "selfless." The "bad" fictionalized dog also possesses many "human-like" qualities; they are spiteful, defiant, willfully disobedient, and unappreciative, especially when failing to cooperate with their human keepers. In actuality, dogs do things for dog reasons; everything they do works in some way for them or they wouldn’t be doing it. Dogs do seem to genuinely like us, and like being near us, provided we make them feel good, and help them experience positive consequences. No, dogs don’t want to unselfishly "please us," and yes, dogs actually are self serving creatures; no apologies, and no hard feelings intended.  For some reason, many people believe dogs should want to do what people suggest, without any concern for what might be in it for themselves.

     Dogs are social creatures, able to cross species-specific boundary lines and hang out with us as surrogate group members; but they are not able to drop all of the dogs things--canine drives, instincts, and needs--that make them dogs (and which also make them so appealing to us!), in order to better fit into our lives. Dogs are not aware of convenience and inconvenience to us; are not aware of dollar value of possessions, or of messes, or the worry they sometimes seem to cause us, or any of the concepts and repercussions of "right" and "wrong." Dogs learn by their own experiences, and increase or reduce behaviors to improve results for themselves. When people live with dogs, and become involved in their dogs’ experiences, they become a part of their dog’s training, whether they intend to or not.

       Everyone who lives with a dog engages in some form of training, even if it is accidental. If an owner walks to the kitchen and opens the fridge, gets some food, and begins to eat, there is a possibility the dog, smelling food, will appear to investigate. The dog has never experienced this before, and has no hope of generosity from the owner, no "dominant" designs upon the owner’s food; it is simply attracted to the smell of food. The dog loving human may share some food. What are the person’s intentions? It doesn’t matter--the dog has just experienced a positive result, obtainment of food--and the likelihood of it approaching the owner again in this situation increases. Learning has occurred, on both sides; for the dog -- "I can get food in this situation," and for the person -- "Food attracts the dog toward me, possessor of the food, in this situation." People learn by the same rules of reinforcement that apply to dogs. Because the owner has been positively reinforced, the probability he or she will repeat the food sharing behavior becomes greater.
 
Dogs (and people!) can be trained to make associations and perform specific behaviors in reaction to consequences:
Operant conditioning is the conditioning, or strengthening, of behaviors using positive consequences. Dogs can be taught to increase their performance of specific types of behaviors, or operants, such as jumping, sitting, barking, or anything else they do, by experiencing a positive consequence, or reinforcement, associated with the behavior. The probability that a dog will repeat an action or response increases as a dog experiences reinforcement. Classical conditioning is the pairing of things which have no meaning with things that have meaning, causing them to become associated. When this association is made, the once meaningless thing becomes meaningful to the dog, causing the same reaction as the meaningful thing did.

Dog training methods usually uses both kinds of conditioning; trainers manipulate consequences to influence dogs’ behaviors, and also train dogs to associate words, signals and cues with specific behaviors and consequences.

While there may seem to be many different training methods, and countless opinions on how training can best be accomplished, dog training systems can be categorized as either punishment and avoidance based, positive reinforcement based, or a combination of both. Because behavior is under the control of its consequences (the law of effect), dogs will always do what works best for them. Dog training is always based upon starting and stopping consequences, either negative or positive, in order to influence a dog’s behavior.  The simple summary of life's motivating forces would be "seek what is perceived or experienced as positive, and avoid what is perceived or experienced as negative." There are four types of consequences which effect behavior: 
Reinforcements are perceived as positive:
positive reinforcement --causes positive things to start
negative reinforcement--causes negative things to end
Punishment is perceived as negative:
positive punishment causes a negative thing to start
      negative punishment causes a positive thing to end

Some training methods focus primarily upon the use of punishment to change a dogs behavior. Punishment can be remote and impersonal, seeming to come directly as a result of the dog’s own impulse, or from something specific in the environment, or punishment can be personal, coming directly from a punishing "disciplinarian." No matter how convinced people are that their dogs "know right from wrong," proven by cowering, supposedly guilty postures, the proven reality is dogs associate personal punishment with the punisher, and with performing specific behaviors in their presence. They do not change their mind about the appropriateness of a behavior because of a supposed "higher up’s" opinion; they learn to suppress behaviors when the punisher is nearby posing a threat. Properly executed remote punishment takes a great deal of sensitivity and timing; done incorrectly it can easily over-inhibit or traumatize a dog. It is never angry, judgmental, intended to harm, or used to teach a dog a behavior is "wrong"; it is used to cause certain behaviors to fail, or be unsuccessful. A dog that has just come to a behavioral road block with one of its own previously rewarding impulses will be more receptive to redirection into an alternative, positive reinforcement-earning behavior provided by the trainer. Remote punishment can be used as a part of a "correction"--to help make a particular behavior work best for the dog. Secondary negative signals can be established, by pairing sounds (pennies in a pop bottle or can) or words (eh!) with punishment. Paired with a remote correction, the sound eventually becomes meaningful by itself, making the actual correction obsolete. When people use their voice to deliver secondary negative reinforcements, they often have a hard time keeping it impersonal. It is supposed to be a sound, like a buzzer, which by association with past punishment situations warns. But paired with anger, eye contact and disapproval, the sound becomes a warning of personal punishment and serves to intimidate, not educate a dog.

Many people still hold onto their firm belief dog’s need personal, disapproving punishment in order to learn. Personal, emotionally-based punitive training seems to occur most often amongst people who have a notion that dogs need "to be shown whose boss," or "taught what they are doing is wrong." People acting as "parents" to dog "children" often personally discipline their four legged "fur-kids." People who have taken an exaggerated perception of canine "pack mentality" into their relationships with pet dogs often attempt to discipline as "alpha" leaders, attempting to use supposedly dog-specific personal punishment to show their subordinates what not to do. Some people perceive dogs as animals possessing an innate desire "to please their masters"; noncompliant dogs are often viewed as willfully defiant rebels. Punishment based trainers frequently voice extreme disgust at the current increase in positive reinforcement training systems, especially those which are heavily food-reinforcement based.

One trainer I interviewed, who asked not to be named, told me he feels currently popular "all positive" methods are incomplete. This trainer, who I will refer to as Steve, has been training dogs in his clients’ homes for twenty years. He believes all-positive methods fail to teach dogs what NOT to do, in a manner that rules inappropriate behaviors out. He described his image of the stereotypical positive reinforcement trainer--"a person using handfuls and pocketfuls of hotdogs and liver to bribe their misbehaving dog, shoveling food into its mouth every twenty seconds." He claims his own clients "agree with his system and do not want to reward bad behaviors; they want to teach their dogs to stop them." Steve trains using spoken commands, followed by "corrections"--a jerk on a choke chain collar--if the dog fails to "obey." Asked why he felt dogs might disobey, he stated "defiance, dominance, and disrespect," adding "when people put them in their place, dogs feel more respectful and secure." Steve gained his training education in the military, working as a K-9 trainer, and later assisting at a kennel that trained protection dogs. His main complaint regarding pet owners failing to properly educate their dogs is "they are too soft on them; they forget they are dogs."

Another trainer I spoke with believes "dogs need to be treated as dogs, and their trainers need to think and act like dogs as well, but superior ones." This trainer, "Marge," believes her methods are very positive. She "gently but firmly" rolls dogs over into submissive postures, pinning them, and then rewards their "submission" with groin petting to establish "dominance" over them. She also uses vocal grunting-- "correction growls"--to "show her dogs her disapproval for certain behaviors, using their own language, like their dam would." Marge is fiercely opposed to the use of any food in training reinforcement, claiming "alpha doesn’t share." She believes dogs do need positive reinforcement, but feels the only positive reinforcement a dog should need is security and approval from "higher ranking pack members, whom dogs naturally seek to please." The biggest problem Marge sees families have with training reinforcement occurs "when dogs learn to respect some people as alpha, but take advantage of weaker ones." She believes "it is a failing on an owner’s part not to assert convincing superior dog force, and that children must be taught from the start not to let the dog push them around." For misbehaviors, Marge advises owners to "catch misbehaviors in the act, and then grunt, roll, pin, forgive ("wolves are forgiving"), and keep trying over and over until the dog gets it right."
A third punishment-based trainer, "Pat," exclusively uses electronic collar corrections to teach dogs avoidance and to "correct command refusals." Pat feels "dogs need to be taught what happens when they refuse to obey." Pat believes in "being discreet with corrections, to avoid scaring the dog, or making a negative association between the owner and dog." Pat adds, "The dog should never know the owner or trainer caused the bad consequence; the dog needs to blame negative experiences on its own actions or failures to respond correctly." To accomplish training with clients’ boarded dogs during their three week stay, Pat sits in a chair in a training area, pretending to read a newspaper "so the dog doesn’t blame me." When the dog "sticks its nose where it shouldn’t"--the area is heavily baited with overflowing wastebaskets, shoes, socks, children's toys, tissues, cat food, etc.--Pat silently gives the dog an electronic collar "correction." Pat claims student dogs become "Velcro," and "very easy to train, except for the command ‘stay‘--the dogs become so attracted to me they can’t stand to stay away!"  Asked how the dogs are taught to stay, Pat explained, "When they break stays they are e-collar corrected and then pulled back to the spot. They know they are doing it wrong, and don’t want to go back, because they really want to be by me; but they can’t have the reward until they get it right." Asked what the reward was, Pat explained "being near me, of course; and my praise too!" According to Pat the biggest difficulties in training are "the owners; they aren’t me!"

The punishment based (whether they realized it or not) trainers I spoke with all seemed to view purely positive training methods, especially those which used food as the primary reinforcer, as "too soft," and believed the approach skirted the issue of "teaching dogs right from wrong," focusing only on "bribes."
I questioned several novice (first time) dog owners to get their perspectives on positive verses punishment based training. While all of them voiced a strong desire to be "positive" in their training, most seemed to think the term "positive reinforcement" was synonymous with the exclusive use of food as the primary reinforcer. Several confided worries about training their dog using (their perceived application of) positive reinforcement, including questions such as: "What if I don’t always have food available?"; "What if my dog wants something else (a squirrel, cat, etc.) more than the food?"; "What if he prefers food, and stops having any interest in my praise and affection?" Other concerns regarding the use of food as a primary positive reinforcer included: disinterest in the food; increased elimination or frequency of elimination in reaction to the food; and increased expectation of food and resulting fixation (to the exclusion of everything else) in the presence of food. A deep rooted perception that the exclusive use of positive reinforcement in training will reduce respect and social connection between people and dogs persists in the minds of many trainers and pet owners. All of the novice owners I spoke with also told me they strongly believe it is necessary to train dogs using the occasional application of "some" personal punishment, thought they wished "they didn‘t have to." The most commonly used or approved of methods for punishing misbehaviors included scolding, removing, and "showing and shaming" dogs. Hitting "lightly on the rump or snout" was also admitted to by several owners, who felt it was their only option based on the "severity of the misbehavior."
Positive reinforcement trainers say not only is it unnecessary to personally punish dogs, it is abusive. They take a firm stance against overly firm, punishment-based training programs, citing cases of abuse, harm and even death being inflicted upon dogs in the name of so-called "training." Listing a veritable smorgasbord of Marquis de Sade-esque devices and methods developed for fighting the battle against "bad behavior," positive trainers describe supposed teaching methods which hurt, confuse, and abuse dogs. I spoke with a trainer who now strongly advocates the use positive reinforcement, but early in her career instructed at a training school which utilized very harsh punishment tactics.
On the condition she was allowed to remain anonymous "Mary" confided a grim inside view of what she refers to as "the wrong way to train a dog":
"I didn’t know there was another way; I was referred to the school with my own dog, who did pretty well, and just sort of worked my way up the ladder until I was an instructor there. Some of it never seemed right, but the end supposedly justified the means. And there were things I knew were wrong, but which we rationalized, or grew insensitive to. There was an inside joke at this school (which is no longer in business, by the way), amongst the instructors, about some dogs having ‘frequent flyer mileage‘ they spent so much time in the air, being jerked around on their choke collars. The feeling was, if a dog didn’t ‘back down’ when confronted by the handler’s force, it was being defiant. Many dogs became defensively aggressive under the trainer’s authority, and were deemed ’untrainable.’ The subject of fear-based defensiveness only came up if a dog looked really pathetic, or terrified, and then they were always suspected to be ‘fear biters.‘ We never really considered dogs had a right to be afraid, that we were what they were afraid of, and that we were creating a need for defense from us. We thought, and wanted the owners to think, dogs shouldn’t be allowed to think for themselves, they should get all of their answers from us, and were not ‘trained’ unless they completely gave in and surrendered to our will. I think I helped make a lot of dogs more messed up, when all I had ever wanted to do was help them."
Stephanie Shain, Director of Outreach for Companion Animals at the Humane Society of the United States, says the organization disapproves of the use of any punishment or aversive devices which cause pain. HSUS does not take a hard line against the use of mild aversives such as scents or sounds to discourage inappropriate behavior; but in order for their use to be humane, they must be used correctly and be appropriate for the individual dog. "It is better to show a dog what it should do, rather than what it should not do," says Shain.
The Association of Pet Dog Trainers (APDT) is a professional organization of individual trainers who are committed to becoming better trainers through education. Founded in 1993 by Dr. Ian Dunbar as a forum for trainers to associate with one another and to discuss topics of mutual interest, APDT has since become a member-driven organization producing educational conferences in locations throughout the United States. Their annual conference has grown from 250 participants a few years ago to 1500 participants in 1999, and their worldwide membership includes 2500 members worldwide.  Members of the Association of Pet Dog Trainers pledge to adhere to the guidelines of the APDT Code of Ethics which include:
4)The practices of hanging, beating, kicking and all similar procedures causing the dog great pain, distress and imminent potential for physical harm are inconsistent with humane dog training. These procedures represent a serious violation of professional ethical conduct and will not be tolerated.
5)When designing training and behavior programs, members should take into account the physical and psychological well being of the individual dog. Positive trainers claim punishment-based approaches treat mans’ best friend more like a crafty enemy than a student needing direction to realize its potential. Positive trainers feel there is never a need to punish a dog for behaving incorrectly; instead they focus their educational energies on teaching and motivating their students to perform appropriately.
In answer to accusations they are creating robotic, food seeking automatons instead of loving companions, positive trainers state "the people who say that don’t understand learning theory." According to Jean Donaldson, in her book The Culture Clash, "the two educational areas where the greatest amount of myth and knowledge void are; 1) dog behavior, i.e. the genetic endowment and constraints or "hard wiring" the dog comes with and 2) animal learning, i.e. the nuts and bolts about how experience affects the behavior of dogs and other animals, including us".
Food-reinforcing "positive" trainers state the food is a temporary tool, used to initially lure and/or reinforce desirable behaviors; as the behavior becomes more automatic, the food is "faded" from frequent use, replaced by more subtle, but equally positive reinforcers. Positive training methods employ food as one of many desirable rewards trainers can use to motivate their dogs. Because food is often desired by a dog, it can work in early training to motivate frequent repetitions of a behavior a dog might otherwise offer less frequently. Using food as a primary positive reinforcer will not cause a dog to suddenly turn into a brainless automaton, or want to rip out its owners throat. Used excessively and improperly, food may reinforce grabby, pushy, expectant behavior, and allow assertive dogs to walk all over their less assertive trainers. Improperly used, food reinforcement can also actually eclipse more subtle positive reinforcers, or reinforcers more appropriate and natural to the situation. Dogs in experiments in laboratories may have proven most willing to work for food, but dogs in a home environment, especially if they are well fed, may not always be motivated by a treat. In fact, a dog interested in chasing a squirrel may well turn its nose up at a big slab of liver, preferring to pursue the "hot lunch."
Anything a particular dog wants in a given moment can work as a positive reinforcer; dogs are constantly seeking and gaining positive reinforcement without human help. When people make what dogs want contingent upon a particular response, they are training using life's positive reinforcers! Allowing a dog to chase the squirrel as a reward for walking away from the squirrel, (preventing it from chasing until it reduced all tension on the leash and followed) would result in a dog looking to and willingly following its owner when interested in a game of team squirrel chase (this also buys the squirrel a bit of time to increase the distance). This example used both negative punishment (the self-obtained good thing ended) and positive reinforcement (the trainer-delivered good thing started), without requiring any anger, scolding, hitting, and without requiring a "handful of hotdogs." When a dog shows interest in play, play is the reward; when a dog is tired, rest is the reward, when a dog is eager to go outside, going out is the reward; when hungry, eating is the reward; when socially attracted, attention is the reward. Waiting until a dog is sitting to open a door reinforces sitting; sitting works, so the dog becomes more likely to offer sitting.
The best trick in positive training is pairing a secondary signal, such as a sound (clicker) or spoken word (good!) with various positive reinforcements. This allows the secondary signal to be used to "mark" desirable behaviors, promising reward at the instant the behavior occurs. With proper reinforcement, dogs want to behave in the ways suggested by their trainers. Positive reinforcement helps strengthen the social bond between human and dog. The positive trainer actually becomes a secondary positive reinforcer, whose very presence promises success and positive experiences!
Professional opinions seem to be growing progressively more unilateral, leaning heavily to the positive reinforcement side and away from punishment. Years ago there were many different training systems, all explaining dog learning and perception in an array of conflicting manners. Based upon subjective interpretations and strong personal opinions, trainers outlined training steps that fit their particular descriptions of dog thinking and intelligence. These days, an increasing number of objective trainers are following the path provided by knowledge of learning theory; it is harder to stray away into dark subjectivity when using a clearly marked map. Trainers eager to accomplish goals with minimal negative fallout are focusing upon positive ways to help pet owners make educated, ethical training choices, rather than teaching ways to suppress undesirable behaviors.
Paulette Allen, a member of the APDT and a positive reinforcement trainer, feels the key to success is paying attention to the most important opinion--the dog’s!
"Positive training trends are a huge step forward in the right direction," says Allen. "However, the training term ‘Handler to dog feedback’ should not be primarily ‘food’ rewards. Excessive focus on quick end results can actually impede long-term training goals. It isn’t enough to just manipulate behaviors; positive trainers need to seek deeper understanding and communication skills that could enhance the individual dog/human relationship. I personally feel more connected to an individual dog, as that I am a better trainer, when utilizing more creative positive motivators, other than food, which can include the dogs own sense of accomplishment! Demonstrating respect while training, by refraining from inflicting harm and providing positive reinforcement that enhances not just a dogs actions, but its perceptions of people, should be our goals as trainers." Teaching better understanding of dogs’ natural behaviors and instinctive drives, and teaching non-combative methods of management, trainers are helping owners accept their pets for what they are. Using methods which guide and reinforce, rather than fight and force, dogs are helped to achieve their personal best behaviors and social relationships. As knowledge increases, the use of pain and punishment to accomplish training goals decreases, replaced by concepts and applications for reinforcing desirable behavior positively. 





THE ASSOCIATION of PET DOG TRAINERS 66 Morris Ave., Ste. 2A, Springfield, NJ 07081
Positive Reinforcement Books:

Behavior Problem Solving in Dogs
William Campbell, 1975, American Veterinary Publishers, Santa Barbara, CA

The Culture Clash
Jean Donaldson, James and Kenneth Publishers, 2140 Shattuck Ave, #2406, Berkley, CA

Don’t Shoot the Dog
Karen Pryor, 1985, Bantam books, New York

Excel-erated Leaning: Explaining in plain English how dogs learn and how best to teach them
Pamela J. Reid, Ph.D., 1996, James and Kenneth Publishers, Oakland, CA

How to Teach a New Dog Old Tricks
Ian Dunbar, Ph.D.,DVM, James and Kenneth Publishers, 2140 Shattuck Ave, #2406, Berkley, CA