Turning Students into Dog Trainers
When I teach a reactive dog class, I tell my students that I am not going to give them cookie-cutter recipes to deal with problem behaviors. Instead, I am teaching them to be real dog trainers. I want them to be able to solve any behavioral issue on their own should they need to in the future. I want them to be analyzers, problem-solvers, and architects of behavior who feel confident they have the understanding of the mechanisms of behavior so they can fix things when they break down. By becoming familiar with the nuts and bolts of learning theory principles, they will be able to think logically and intelligently about how and why their dogs act in the ways that they do.
No More “Stupid” or “Stubborn” Dogs
There is such a huge difference between my reactive dog students and the beginner obedience students I had years ago. In my beginner obedience class, if the dog misbehaved, the student complained that the dog was “stupid” or “stubborn.” If the technique didn’t work, it was a problem with the dog. It was a “do-as-I-say” era, where few students questioned the teacher or asked “Why?” when the instructor presented a technique, regardless of how uncomfortable they may have been with the instructions or the lack of results they saw with their efforts. It was only after I took competition obedience classes with Patty Ruzzo that I began to ask questions.
Previously, I had simply handed over my leash. I was the student, and the teacher was the trainer. Who was I to question authority? This was the trainer; she must know what was best for my dog, right? But what if my dog disagreed?
Patty explained to me that if I wanted to be able to communicate effectively with my dog, then I had to develop the skills necessary to “read my dog” and recognize what his behavior was expressing. How else could he count on me to make the right behavioral choices for him? Learning to question and to listen to my dog when I made choices that affected his welfare was a great gift Patty gave to me and is one I endeavor to pass on to all of my own students. The dog is the teacher. When students know how to read their dogs, when they know how the mechanisms of consequences drive behaviors, they are better guides, advocates, and friends for their dogs. My students know how to interact with their dogs in a meaningful way, a mutual flow of information passing between dog and human, and, therefore, know how to set themselves and their dogs up for success in any situation.
In my reactive dog class, if a dog is struggling, it’s not a “dog problem,” and the dog is not being stubborn or stupid. It’s a kink in the chain of our training plan and a red flag indicating that we need to reevaluate and make changes. The students learn to ask themselves questions such as, “Was the exposure to the trigger too long? Too close? Did I ask for too much too soon, and if so, how can we get back on track?” They are not complaining about the dog; they are behavior sleuths trying to solve the mystery of what went wrong.
The learning theory principles that I employ are based on Karen Pryor’s book Don’t Shoot the Dog. This is a book on behavioral theory firmly grounded in science but written in a manner that is entertaining and accessible for laypeople interested in behavior at any level. Don’t Shoot the Dog is not a dog-training book; Karen Pryor originally wrote it for humans, especially teachers, coaches, and parents. Fortunately for dogs and their people, this book has had a trickle-down effect, and those committed to training techniques based on positive reinforcement have adopted its principles, popularly embodied in “clicker training.”
Below are the principles that I’ve found are important for handlers of reactive dogs to learn.
The Formula for Learning
Based on the principles of operant conditioning, clicker training is a positive-reinforcement training system that incorporates the use of a marker signal (the click) to tell the animal precisely what it is doing right at precisely that point in time. Training progresses because behaviors that are marked by a click and followed by reinforcement are more likely to happen again. Those behaviors “pay.” Because of its effectiveness in building behaviors, clicker training is particularly suited for training new behaviors. The premise of clicker training is quite simple: you see the behavior, you mark it with a click, and you reward the behavior. For more on the benefits of clicker training, please see here.
Some students initially reject the idea of using food as reinforcement, fearing that their dogs will grow dependent on it. For such students, examining the differences between luring and rewarding behavior can be extremely helpful in clarifying when and how food reinforcement can benefit training and dispelling the idea that food-trained dogs won’t perform without it.
When you lure a behavior, you present the reward before the dog offers or completes the desired behavior. You offer the dog the opportunity to decide whether the lure is worth the effort; in essence, the dog performs an on-the-spot cost-benefit analysis that happens so quickly it would make an accountant’s head spin. Take, for example, the owner who lets her dog out first thing in the morning for a potty break. Once the dog is finished, she calls him back to the house. The dog hears his owner calling him but continues to sniff the grass. The owner calls again, this time louder and, perhaps, followed by more colorful language when he doesn’t respond. Gleefully, the dog buries his nose further into the grass, flopping on his back to roll in whatever stink has caught his attention.
The owner resorts to more drastic measures. She heads to the refrigerator to retrieve bits of last night’s roast beef. Showing the dog the treat in her hand, she calls again. Happily, the dog runs into the house to get the roast beef.
This technique is well and good if two requirements are met: the handler is always willing to pull out a lure (and because dogs can get bored with a single reinforcer, she will have to mix it up frequently), and the lure she offers is always better than the environmental rewards the dog can give himself by ignoring her. In this way, luring can create “Show-me-the-money” dogs reminiscent of Cuba Gooding, Jr., in the popular movie Jerry Maguire.
What is the dog really learning here? That he can quickly train his owner to continually offer better and better treats for lackluster behaviors. If the owner of the dog above first offered a bit of chicken and he ignored it, she would grab the roast beef. If that didn’t work, she would upgrade to the filet mignon. Soon, the dog would not work for anything less than authentic, grass-fed Kobe beef, cooked to medium-rare perfection. While these dogs are frequently called stubborn or stupid, they are actually smart—with a great talent for training their humans.
Whether or not skeptics of using food rewards are right when they claim that lured dogs can become dependent on food, what is true is that lured dogs often learn new behaviors more slowly and less completely because they tend to concentrate on the reward instead of on what they are doing for the reward. When you use food as a reinforcer for a completed behavior, however, you get a different result: Instead of the dog making the decision about whether the treat dangling in front of his nose is worth performing the cued behavior, the trainer retains control of the situation. The dog has to work (offer a desired behavior) to get what he wants. He also has to buy into the system enough to trust that if he responds to a cue, he’ll get paid—in treats, praise, play, a run in the park, or something else he wants. It’s a win–win situation: The owner wins because her dog does what she wants; the dog wins because he gets what he wants. A consistent trainer who follows learning theory principles uses food rewards effectively; the dog thinks about what he needs to do to get the reward and learns faster and better, without needing visible treats to perform.
That said, there are some benefits of and appropriate uses for luring. Think of luring like a hot fudge sundae: fine occasionally but not something that is part of your daily diet. Use luring in a pinch but not for the long-term. It is best suited for jump-starting behaviors. Use luring in the short-term and always with a plan in mind of how to get the lure out of the behavioral picture as soon as possible.
For instance, I use luring to help a dog develop muscle memory for a given behavior. Years ago, I was training one of my dogs for the sport of Canine Freestyle. One of his moves was to heel alongside me while we moved laterally together. Because he had never moved this way, I stood beside him and glued a treat to his nose. Very slowly, I moved the treat in a straight line away from me. As my dog was learning to move in this way, he was stepping on my feet and tripping over his own! It only took a few repetitions for him to realize where all of his feet needed to land to complete his moves. Quickly we transitioned to “true learning”: one correct step is taken, click, and treat in the desired position.
Whenever you are training new behaviors, it is best that you start in a neutral, distraction-free environment. This allows the dog to offer his full attention to the behavior at hand without having to divide his attention between you and the learning environment. Just as humans like a quiet place to study, dogs require their own “learning sanctuaries.” This is especially important for rehabilitating dogs with fear and/or aggression issues. When any dog is startled, his ability to learn decreases and his propensity to react dangerously increases.
Advise your students to start working in a quiet room of their home with no distractions, and then gradually add distractions to the learning environment at a pace dictated by their rate of reinforcement, an objective measurement of their dog’s success at any given stage in the training process. Once the dog is able to perform effectively with a variety of distractions in the house, the student may choose to take the behavior to the backyard when no distractions are present, and then gradually increase the amount of distraction outdoors. Students can use similar protocols when taking these behaviors on the road for continued opportunities to generalize behaviors.
The key to adding distractions successfully is to do so slowly. While it may seem counterintuitive, the more a student lets the dog determine the pace at which the learning proceeds, even if sometimes it feels as if it is at a snail’s pace, the faster the dog will learn.
One Criterion at a Time
When developing a behavior, it is important to work on one piece at a time. If a student is teaching her dog to hold a sit-stay, she will need to develop the sit behavior itself first, and then work separately on distance, distractions, and duration before she begins to combine any of those criteria. The dog can only learn one piece at a time reliably. I always warn my students not to be surprised if, when working on one criterion, other pieces of the behavior temporarily degrade. Once the dog has learned the new piece, only then can the trainer begin combining criteria.
Lower the Criteria
When we increase the level of distractions in the environment, expect the behavior to deteriorate in some form: the dog’s response may be slower, more hesitant, and briefer in duration. Deterioration of parts of the behavior is a necessary—and natural—part of learning.
We all have experienced a dog that lies down reliably in response to a verbal cue at home but, when out for a walk, ignores the cue when his handler stops for a chat with the neighbors. A handler who doesn’t recognize the need to adjust criteria relative to the dog’s previous level of training and the current environment may be tempted to see her dog as stubborn and feel frustrated, angry, or embarrassed by her dog’s inability to respond to the down cue in this more challenging situation.
Criteria adjustment is an important concept for students because it helps them avoid getting discouraged thinking that the dog has forgotten behaviors they worked so hard to develop. Because of students’ own learning histories, emotional perceptions of the dog’s behavior, and previous training frustration, students in a reactive dog class are particularly vulnerable to being disappointed when behaviors seem to fall apart.
Use Appropriate Reinforcement
I remind my students to use reinforcement that is equal to the level of distractions in the training environment. This is not only important in selecting reinforcements but also in deciding how fast to dispense them (the rate of reinforcement).
As stated in the “Student Equipment” section ( here), ask your students to number their dog’s favorite treats from 1 to 10 (with 1 being the lowest-value treat and 10 the highest). Use #1 treats (Cheerios, pieces of kibble, and so on) in quieter environments like the house, and reserve the #10 treats (steak, peanut butter) for the most challenging behaviors or environments like training class, the vet’s office, and so on.
Students also need to calibrate the rate of reinforcement to the challenges of the environment, and they need to develop the mechanical skills to vary that rate (See “The ‘Feed the Cup’ Game,” here). Especially when teaching a new behavior or operating in a challenging environment, aim for a high rate of reinforcement.
Ben and I once attended an experimental class in training clicker techniques that Karen Pryor offered in Massachusetts. One night, Karen asked Ben and me to heel past all of the dogs in class while maintaining a high rate of reinforcement. I followed the instructions as I understood them, feeding Ben approximately every two to three steps.
Karen then asked me to push that rate of reinforcement even higher. She gave me a measurable goal: “I want you to feed him 50 times as you move from here to the wall.” I looked at her in surprise, “I don’t think I can feed him that fast!”
“Try,” she said.
So I did. I clicked and treated Ben for every single step, moving him down the length of the room. The difference in Ben’s behavior was remarkable! (He probably thought the same about me!) He heeled beautifully with me, eyes sparkling, tongue lolling out of his mouth happily, not once breaking eye contact with me to notice the other dogs in the classroom. We were in a bubble of reinforcement, so in tune with the process and each other that the distractions had ceased to be distractions anymore!
So, just as a student should lower criteria when facing challenges, it helps her dog be successful if she raises the rate of reinforcement.
The Training Session
Training sessions should be short in duration and high in fun. Before starting each session, the student should decide which behavior she is working on and which criterion she will be clicking and treating. While two to three training sessions per day are ideal, one or two a day can still result in dramatic improvements in her dog’s behavior and in their relationship.
During a session, the student should focus solely on the dog. Put the cell phone away, busy the kids with an activity, close the door, and put a “Do Not Disturb” sign on it if necessary.
Take breaks between sessions to give the dog the opportunity to rest, play, or get a drink.
The Training Process
When training new behaviors, it is best if a single handler in the family takes responsibility for training any given behavior until that behavior is reliably on cue. Multiple handlers in the early stages of training new behaviors can hinder the learning process because of inevitable inconsistencies in criteria or in delivering reinforcements.
There are a million ways up the mountain and nearly as many ways to train new behaviors. Too much information can be confusing to novice handlers; I’ve found it works best to give the student a single set of instructions to begin with and encourage her to follow them as closely as possible. If for some reason the student does not understand the directions or the instructions are not producing the intended results, the trainer can provide the student an alternative teaching strategy.
Once the handler has trained a behavior to the point where it is on cue and the dog responds happily and reliably to the cue, the handler can then instruct other family members about how to cue the dog for the behavior and reinforce as needed.
To test the behavior, my rule of thumb is to ask for the behavior three times. If the dog can respond reliably, rapidly, and confidently each time, the handler can begin to “take that behavior on the road” by introducing new distractions to the learning environment. When the dog’s responses do not meet criteria for the three-trial test, the handler needs to evaluate which piece of the behavior requires refinement and then establish a training plan to patch that hole in the learning process.
Reinforcement, especially when it involves play, should be part of the training session, not a “reward” once the training session is over. A student should signal the end of the session with a neutral verbal cue like, “All done” or “No more.” We want the dog to be a little disappointed at the end of a session. I have seen far too many students work their dogs, conclude a session by saying, “OK!” and then play wildly with their dogs. Such play makes the end of a session more exciting than the session itself and dampens a dog’s enthusiasm for learning. If the handler had inserted that much fun into the session itself, she would have a dog that couldn’t wait to work again because working is so much fun!
Leadership without Confrontation:
Effective Home Management
Rehabilitating reactive dogs often requires retooling the relationship the owner has with her dog. Reactive dogs need leadership based on mutual love, respect, and support. I never use intimidating or threatening techniques. Dogs are quite willing to relax into a structure that is created for them out of a place of patience, fairness, and consistency. The clicker home-management program that I use in my class helps students show this type of leadership to their dogs within their homes.
Frequently you will find that dogs have trained their humans far better than the humans have trained their dogs. I have been in many households where the dog literally controls the behavior of his human family members: the humans will do whatever the dog wants when he wants them to because, if they don’t, he will bite them.
Years ago, I visited a family who had an adorable four-month-old wheaten terrier. A cuter dog would be hard to find—this one looked like a Gund teddy bear! I was called to the home because the dog was guarding his crate and bit anyone who walked by it, regardless of age, gender, or level of familiarity with the dog. This was anything but puppy biting! The parents, not wanting to teach their children that dogs are disposable, were loathe to return the dog to the breeder.
Far from thinking his behavior was a problem, this young wheaten viewed the humans in the house as the ones who were struggling with behavior problems, and that the solution was to “correct” those inappropriate behaviors through using his teeth. His behavior caused ripples of change in the behavior of his entire human family. The parents established new traffic patterns, re-routing the children through another door so that they would not go near any of the puppy’s things (including the crate). The parents conducted a “mood assessment” before they approached the terrier. These patterns began to permeate nearly every aspect of this family’s life.
I instructed the family to begin using the exercises described in my book, Click to Calm, especially ignoring demanding behavior and asking the puppy to respond to a cue before giving him what he wanted, like attention. Before we even started having regular clicker training sessions, the parents saw a dramatic reduction of undesirable behavior in their wheaten puppy and an equally dramatic improvement in their own relationships with the dog. They began to feel optimistic about his prognosis. The family was reclaiming their life from the dog, and, as a benefit, could more fully integrate the dog into the family life that they had desired all along.
I like giving these principles to my reactive dog students because they can only help the situation at home, regardless of the current relationship each student shares with her dog. Each week, I give my students one new home-management principle. It is imperative that students make these changes slowly and that they understand the importance of each exercise.
Change is hard. If a dog has been rehearsing and being reinforced for undesirable behaviors for years, he may initially resist some of the changes in the protocol. Additionally, students will have well-established behavior patterns that help maintain these unwanted behaviors in their dogs, so the people need to be re-trained, too. Learning never happens in a straight line—there are always peaks and valleys. To avoid confusion and frustration, warn your students about extinction bursts, since unwanted behaviors may temporarily get worse before they get much, much better. Give your students tools and strategies in advance so that they are well prepared to deal with such inevitabilities as they occur. When students have a plan and are able to meet challenges with instruction rather than with emotional responses, the change will roll right past any speed bumps on the path to better behavior.
The Case against Positive Punishment
I am very clear with my students up front about using positive training techniques to modify their dogs’ behavior, and I review with them carefully the following possible negative consequences of punishing their dogs:
1. Regardless of the technique one uses to modify behavior, it is best to implement a single method to keep the training process consistent for the dog. If you positively reinforce a dog in one situation and heavily punish him in another, you end up strengthening his belief that the world and the environment are unpredictable and potentially dangerous, exacerbating his potential to bite.
I have had students in the past who, midway through the class, inform me that their dog’s behavior is not improving but deteriorating. This immediately raises a red flag because, in nearly every one of these cases, the student is combining the punishment-based methods of her past with the newer, more positive methods she is learning in class. Rather than instilling a sense of safety in the dog, this mish-mash teaches him that his environment is even less predictable than it was before.
Because living with a reactive dog can be so emotionally (and sometimes physically!) challenging, and because it is human nature, don’t be surprised that owners of reactive dogs often seek out methods that promise a quick fix. Social pressure may make the owner of a reactive dog feel obligated to “correct” her dog for inappropriate reactions, particularly in public, so that her neighbors are aware that she does not encourage or accept such behaviors from her dog. Punishing the dog does little to improve long-term behavior and, in fact, may worsen existing problems, but in the moment the temporary suppression of behavior reinforces the owner.
2. Punishment can quickly become abusive if it is used as the primary method of communication. In the initial stages, the owner may need a small correction to change the behavior. The more a dog is punished, however, the more frequently and severely he may need to be punished to discourage further repetition of the undesirable behavior; this process is known as establishing a “punishment callus.” In such situations, the frequency and intensity of punishment can increase quickly and seemingly without limit.
3. Additionally, punishment doesn’t teach the dog what the handler wants him to do instead of the undesirable behavior. It leaves the dog in a behavioral vacuum where the target behavior once was but with no idea of what the owner would like to replace that behavior with. For instance, you can punish a dog for jumping on people, but without teaching him a better greeting behavior, he is left to choose an alternative for himself. Rarely does the dog select a desirable appropriate behavior without adequate coaching, so the dog that once jumped becomes the dog that now humps, or barks, or mouths incessantly. The handler may be seeking to eliminate the goal behavior but might not be any more impressed by the alternatives the dog offers.
4. When dogs are punished, they become cautious and fearful. This state of mind is not conducive to learning; it is neither calming nor confidence building—the goals that we are trying to achieve. In fact, ultimately punishment achieves the opposite effect, halting the learning process entirely.
5. Punishment-based training techniques damage the relationship between the dog and the handler. Rather than trusting his owner to provide benevolent leadership, the dog becomes wary of the threat she presents. Because his person’s behavior is unpredictable and she has shown herself unable or incapable of keeping the dog safe, the dog assumes the role of guardian—always on guard, hoping to scare the threat away through reactions before the handler begins to react negatively toward the dog.
6.Finally, as Steven R. Lindsay describes in his Handbook of Applied Dog Behavior and Training, Volume One: Adaptation and Learning, positive punishment carries with it the risk of certain side effects, including hyper-vigilance, exacerbated fear responses, impulsive or explosive behavior, hyperactivity, social avoidance, aggression with minimal provocation, loss of sensitivity, and depressed mood.
I make these facts very clear to my students, stressing that, to learn, their dogs need a safe learning environment and clarity. I beg my students, at least for the six weeks we have together, to give the Click to Calm training techniques in their totality a real shot at making a huge difference in the quality of life they are able to share and enjoy with their dogs. What have they got to lose?
Minimizing the Rehearsal of Unwanted Behavior
It is critical to teach students to take away all of the opportunities for their dogs to rehearse undesirable behaviors. Anything and everything dogs practice they get better at, including aggressive or reactive behaviors. This results in an increase in both the intensity and frequency of unwanted behaviors, something all of your students are certainly trying to avoid. This is especially true of self-reinforcing behaviors that dogs engage in just because they are fun and feel good to the dog, from excessive mounting to counter-surfing to barking at the mail lady or the neighbor’s dog if it results in chasing “the intruder” away.
For many of your students, minimizing opportunities to rehearse undesirable behaviors means that they will have to make substantial, though often temporary, changes to the home environment. If the living room windows offer the dog unlimited opportunity to “patrol” all day, hypervigilant as he awaits his next chance to bark at passersby, the student will need to relocate her dog to another area of the house or block his visual access. If the dog runs along the fence line, barking at the children while they play, the student must bring her dog inside when the neighbor’s children are outside.
If a student’s hypervigilant dog habitually jumps on the couch to “bark away” passersby, she will have to block his access to the couch and the window so he doesn’t get a chance to perfect this behavior.
This is a major point to emphasize to your students! Many years ago, I used to mention this principle at orientation and then assume I had communicated the point; I failed to follow up for the remainder of the session. What I found is that a few weeks into the course, students approached me after class to ask, “Does that also mean that I cannot allow my dog to fence-fight with the dogs next door? He does that all the time and it can go on for hours!” After explaining the necessity of stopping the behavior to individual students so many times, I realized it would be more efficient to make mention of this aspect of learning frequently in the context of the group classes, offering different examples of ways dogs might rehearse unwanted behaviors and how students could manage environments to prevent those rehearsals.
Some students may find management in such situations to be little more than a band-aid solution and that the dog never truly learns that the behavior is unacceptable or how to behave differently in the presence of the trigger. Remember, the first step in stopping the behavior is stopping the behavior. That means stopping the dog from practicing it while a suitable replacement behavior is being developed. This is a necessary ingredient in the recipe for curing reactivity.
I tell my students that behavior works like a bank account. Ideally, you make far more deposits than you make withdrawals. You may have to make an occasional withdrawal because of a huge and/or unexpected expense; the amount of financial damage from such a withdrawal is determined by how much you have managed to increase the account through deposits. If you’ve done your work, there is still plenty of residual cash to allow you to weather any potential financial storms that come your way. Reinforcement history for both desired and undesired dog behaviors functions in the same way.
I say to my students, “The more you expose your dog to his triggers successfully without offering your dog the opportunity to rehearse unwanted behaviors, the more you make huge deposits into your trust account with your dog. On the other hand, each time the dog rehearses the behavior, you take out an even larger withdrawal. Are you making more deposits or more withdrawals? If you do a little bit of both, it won’t take long until your account balance will return to zero, which is exactly the amount of lasting behavior change you can expect.”
Every deposit a student makes into her dog’s “trust account” is a step toward building her dog’s tolerance level to all the difficult circumstances she may encounter as she escorts her reactive dog through everyday life.
Setting up and running a reactive dog class successfully takes more care and forethought than the average basic manners class. The students are entrusting their safety, and their dogs’ safety, to you. In Part II, we’ll cover the preparations you need to make and the logistics of managing a reactive dog class.