CLICK AND CROSSTRAIN FOR VERSATILITY
by Ruth I. Kellogg
(2003)
Many of us have read about or seen those truly versatile dogs. You know the ones I mean – the dogs whose certified accomplishments read like an alphabet soup before and after their names. So many abbreviations that the dog’s name is almost lost! Have you not marveled, as I have, on the dedication of the owners in the education of their dogs and in their accomplishments? Is it possible to garner our own “alphabet soup” of accomplishments with our free-spirited, independent thinking, comedic, and intelligent Alaskan Malamutes and still have them happily doing what we ask? I believe it is.
The Alaskan Malamute is an intelligent dog who thinks, reasons, questions, and enjoys laughter. It is not a dog who is so very serious about working that they become obsessed with the task (like a border collie) or have a very serious demeanor (like some other members of the working group). The Mal loves to work – but if it isn’t enjoyable, the desire to work diminishes or disappears. This independent breed is also rather mercenary. They often ask, “Why, what’s in it for me?” If the task they are “working on” is fun and especially if they know that there is a good “pay off” for it, then the work isn’t “work” in their eyes but a game to enjoy.
While there are many tools, aids, and methods that trainers must sort through, when it comes to teaching the Alaskan Malamute, teaching using primarily the positive reinforcement methods of operant conditioning – specifically with “clicker” training is particularly useful. Add to clicker training the concepts of foundation building, crosstraining, and shaping behaviors, the opportunity for the dog to become truly versatile becomes a possibility – and then a reality.
Foundation Building
Before a trainer starts any education of a dog, an idea of what is to be accomplished is imperative. I’m not referring to short term goals of, for instance, getting the puppy house-trained with good house manners but the far away and long term goals and dreams for that particular dog. If an “alphabet soup” of accomplishments is the goal, then the trainer must be fully aware of the importance of laying down a solid foundation of positively learned (simple) tasks that are then built upon and ultimately shaped into more complex behaviors. For example, in order for an obedience dog or a conformation dog to compete successfully the dog must have a solid foundation of what “stand” means. It doesn’t mean statue-like with a particular “look” when a stranger examines. No, it’s just getting from whatever position the pup was in to standing, however momentarily, with all four paws on the ground and the back parallel to the ground. (The easiest way to teach the stand is from a sit position.) The particular look and rock still stance while the handler is standing away from the dog and a stranger examines the dog is built upon or shaped from the simplest task.
The foundation for a wide variety of tasks is actually laid down in the first six months of a puppy’s life. These are simple tasks that, I believe, all dogs should be taught. They are: sit, down, stand, leave it, give, watch me, and walking on a leash. Of course, learning where the pup is in the pack order (with the handler hopefully earning respect of the pup) is also very important. Socialization to a variety of places, people, and other animals is extremely important. I know how difficult it is when not living close to an urban setting to socialize. However, it is much easier to socialize a small puppy than to do remedial socialization with an adolescent!
Crosstraining
I use the term crosstraining to mean the teaching of basic cues in foundation building that can be used in a variety of different activities. If we look closely at a variety of activities we will see that there are cues or behaviors learned that are common to tow or more activities. A simple example is “down.” The most obvious application is in formal obedience. But it is also used, to name a few activities, in agility (on the pause table or in the pause box), tracking (at the start of the track), harness work (handy during breaks), house manners, grooming, vet examinations, hiking, and the foundation for other tricks (crawl, rollover, “bang”). Learning directional cues is obviously an asset for harness work, but it is also used in agility and certainly makes pleasure walks, hiking, and biking more enjoyable.
A dog who has a variety of learned cues can enjoy quite a mixed bag of activities seamlessly. Dogs, unlike humans, don’t differentiate behaviors or tasks as conformation, obedience, agility, and so on. They are just cues for doing something different that will bring positive results. Malamutes hate drilling or constant repetition of exercises that they know. Boredom with doing the same old thing is lessened dramatically by mixing up cues in a formal teaching session. This keeps the fun alive for dogs and trainers. There is no written rule that says a dog who will only see the inside of a conformation ring can’t learn some fun things like tricks, agility skills or harness work!
Shaping
B. F. Skinner defined shaping as the method of achieving progressive changes of behaviors. The trainer has a final picture or goal in mind of what the behavior that being shaped will look like. Using consistent applications of positive reinforcement when the behavior is altered to become closer to the ideal coupled with lack of positive reinforcement at no change of behavior or regression of behavior, the ideal;/goal is reached without stress and with understanding.
Shaping occurs in foundation building and then in building upon that foundation. To return to my previous example of “stand,” when a pup just starts to stand following a lure, he is reinforced. When he stands following a lure on cue, he is reinforced. When he stands on cue, he is reinforced. The foundation is laid as the pup understands the verbal and/or nonverbal cue for getting into the stand position. Then the behavior is shaped slowly and surely into, as an example, the rock solid free-stacked conformation pose with the handler at the end of the lead and the dog being examined by a stranger.
If a trainer is consistent in rewarding behaviors that are progressing toward the ideal, the dog will begin to understand what is desired. But even the most consistent trainer will encounter the dog’s learning plateau. This is what I call the time after a rapid learning phase in which there is no progress at all. In fact, it seems like the dog is going backwards at times! It is a time when the dog processes the information from the rapid learning phase and in a series of trial and error attempts will try to sort out just exactly what the trainer wanted. It is a very frustrating time for the trainer. But it does pass. Then “all of a sudden,” the dog seems to miraculously remember all the previous work and starts to learn (rapidly) again….before the next learning plateau.
When a dog reaches that point, I try to ignore what the dog doesn’t comprehend in the latest “new exercises” and return to the more basic and older lessons that he know well. I always finish a session on a positive note and then let the dog “process” the information. A few days – or more- of not doing the formal training can be extremely beneficial. I’m always delighted at the high level of retention that Malamutes have no matter how long the hiatus between lessons has been.
Shaping can also be applied to a complex set of cues that are combined together. While it may appear to be just one “exercise,” such a combination of cues – correctly termed a behavior chain – can always be broken down to the smallest level, the foundations. It is important to understand that all individual elements of the behavior chain be separately and thoroughly taught and understood before the elements are combined. Failure to do just one of the elements in the chain results in the chain breaking down.
The secret to shaping a solid behavior chain is to start with the last behavior in the chain first which is always followed by a positive reinforcement. Then each learned behavior is chained and linked to the previous behavior back to the foundation. The reasoning behind this is that the dog knows the positive reinforcement comes after the last element (or foundation) of the complex chain. This keeps anticipation of reward in the middle of the chain low so that the dog’s concentration is focused on the total behavior chain. There is no doubt in the dog’s mind when he’ll be rewarded (and should after each long chain).
Operant conditioning with shaping is utilized by talented trainers who teach dogs the skills to assist disadvantaged people. Size doesn’t seem to matter. I recall seeing (on television) a Papillion who, as part of his many duties as an assistance dog, helped with the laundry. He popped into the dryer and pulled out the clothes so the owner could fold them!
Dogs who excel in the film industry are highly skilled in learning different behavior chains. One advertisement I marvel at has a pointer who has mastered spraying a spray bottle! This dog (in two different commercials), takes the sprayer to various locations, sprays, then goes to another location, sprays…
Is there a limit to what the dog can do? The only limitations are the trainer’s imagination, skill in teaching, and of course, the dog’s physical abilities.
Reinforcement schedules
In positive operant conditioning, the dog is motivated to change its behavior in order to receive a reward. I will stress that reward is not synonymous with food. A reward may be a simple “good boy,” pat on the head, play with a favorite toy, a game of fetch as well as food treats. Not all dogs work for food - yes, even some Malamutes! The trainer must find what each dog loves to us as an incentive or answer to “why should I do that?”
When this incentive is ascertained, then the trainer should use it intelligently. For example, if the dog is a chowhound, feeding tiny pieces of the goodie will keep the dog’s interest and appetite up’ if the dog lives to chase the ball, then a short toss or short catch is the equivalent to a small tidbit. Letting the dog have a good long chase for each positive reinforcement will tire the dog and put the emphasis on the play and thus decrease the dog’s attention and desire to learn.
There are three kinds of reinforcement schedules: constant reinforcement, fixed-ration reinforcement, and variable replacement schedules.
Constant reinforcement means that every time the dog changes its behavior in a positive way, it is rewarded. Constant reinforcement is used in the learning phase as the reward reinforces that the changed behavior is what was desired. As the behavior is learned, the reinforcement schedule for that behavior should be changed to a variable one. The other times for the constant reinforcement schedule are each and every time the dog is asked to make a choice, solve a puzzle or after a long behavior chain.
The second reinforcement schedule is termed fixed ratio. There are a fixed number of responses to number of rewards. For example, the dog must come to the trainer three times in a row on cue before being rewarded.
The third reinforcement schedule is variable reinforcement. It allows great flexibility and spontaneity for rewards. There is no set time when the reward will come… maybe after the first, third or fifth response, maybe each time…The dog is left guessing. So “just in case” the reward might be this time, the behavior desired is performed correctly. The classic example for people is the popularity of slot machines or lotteries.
Clicker Work
Clicker work is the popularized term for positive operant conditioning using a conditioned reinforcer. A conditioned reinforcer is something that is totally different from what is in the dog’s daily life. The commercially available metal clicker in the plastic rectangular box is the most common conditioned reinforcer used. But anything unique to the dog can also be used such as a pen light (for deaf dogs) or any type of easily held noise maker such as a hair clip.)
In clicker training, the dog is taught that the initially meaningless signal or sound (e.g. loud click) is connected/ paired with something the dog does want. Simply: click = treat = something he wants. With solidifying of the pairing, the equation is: click + treat = something he wants. As the education progresses, the equation becomes click = something he wants. Ultimately, this something he wants becomes the desire for approval, affection, and his own satisfaction.
The click is a loud and simple signal for “YES.” In the beginning, the click will signal the end of the behavior cued for. As the dog progresses in its education and is on the variable reinforcement schedule, the click can also become a non-verbal “yes, keep going” during a behavior chain. For example, when I teach directional signals to my dog, I use my agility cross-over. My dogs all know and are comfortable with the equipment so walking up and down the planks is not stressful. In teaching the commands, the dog first follows the target stick. As soon as the turn is started, he is clicked and treated (yes, I even stop dogs on the “going down” plank). When the dog reaches the ground, he is clicked and treated. As the dog understands, the treat first disappears after the turn, the target stick’s usage diminishes, and I move farther away. The click immediately after the turn remains until the dog really understands. The dog is clicked and treated only at the end of the full behavior chain. (The chain is going up the plank, turning correctly on cue, and going down the other plank.) Finally, the full behavior chain is put on a variable reinforcement schedule.
Target Stick
The target stick is essentially a wand that acts as an extension of the trainer’s arm. There are some nice commercial ones, but the target stick that I made from the end of an old fishing rod with a nylon tag tied on one end has been very useful, light, and easy to hold and use.
Until I started using the target stick, I couldn’t quite comprehend how useful a tool it can be. The first exercise I “formally” teach my dogs in clicker training is to “touch” the end of the stick. This really helps in their understanding of change behavior (e.g. move), hear the click, and get a treat. Conversely, they quickly learn that no touch = no click = no treat. As an extension of the trainer’s arm, it also allows the dog’s attention to become more focused on the end of the target stick and less on the hand (that may have food in it). For dogs who become obsessed and/or over excited with food treats, this really does make a difference. In experimenting with the target stick, I’ve found it useful in helping dogs to learn to “hit” the contact areas of the agility equipment, point to where I want them (e.g. jump through a hoop), sitting in front or in heel position, moving in heel position, and as a “mark” to go to in distance work.
Puppies and clickers
Puppies can start in clicker training when they are still in the whelping box. When my last litter was born in 2001, I was interested to see for myself how effective clicker training would be for very young puppies. As an introduction to the distinct sound of the clicker and to start the association /pairing of a unique sound to pleasure/reward, as soon as the pups started on puppy food (at 16 days), I started to “click” a couple of times before the dish was placed in the whelping box. The next day, I added, “puppies-puppies-puppies” followed by clicks and then dinner. By the second day, when the litter heard “puppies,” they moved toward the front of the box hollering as they knew dinner was coming! I’ll emphasize, that the pups were all of 18 days old at that point.
At five weeks of age, the pups began individual lessons. Each pup was positioned on my grooming board and was lured into sit and down positions in the beginning. They were clicked the moment they started to move which quickly changed to a click (and treat, of course) when the cued for behavior was done. As with any litter, some pups ere a little quicker to understand, but all responded beautifully the following day when the lesson was repeated.
At six weeks, I changed the criteria and started asking for “group sits” in the whelping box. The first couple of times, individual pups sat and were clicked and treated immediately. The third time, five of the six pups sat on command on the first cue for “sit.” The next time, all six pups would sit as a group on cue, hear the click, and (mostly) stayed sitting until they got their individual biscuits. I will admit, the biscuits were handed out very quickly! Before each puppy left my kennel, the new owners were shown their puppy sitting, lying down, and standing on cue for click and treats. I then taught each pup “leave it” in front of the new owner which served three purposes. First, the owners were able to see how the click and treat aids learning, how to perform this exercise (an extremely important one in my mind), and finally how bright their new pup was!
As I had two of the pups for 6 ½ months until Ananda went to his new home, both he and sister Zhouma continued with clicker training. Sadly, the new owners generally didn’t take advantage of the excellent teaching tool the youngsters already knew. Ananda’s owner has continued to use clicker training on occasion (especially when he learns new tricks). Zhouma certainly has continued her education in conformation, obedience, and tricks – all aided with clicker training.
Applying Clicker Training
What I have founding clicker training is that the dogs’ attitudes undergo a shift into an eager to learn mode. The need for (boring) repetitive work to “teach” an exercise is eliminated as exercises are approached positively and are more like games that the dogs control. In my experience, my dogs have learned more new behaviors in a shorter period of time than I was able to teach using other methods. My “formal” teaching sessions with my dogs is often only 15 minutes long/dog. But during that time, they are exposed to a variety of different lessons with different activities in mind. For example, we’ll do a little work with a contact obstacle (cross-over) which touches on agility in doing the obstacle in control and hitting the contact points as well as emphasizing sledding directions (gee, haw, wait, on-by). Some obedience exercises are done with heeling, stays, comes, and fronts. Some “fun things” such as different tricks are woven in among the more “serious” exercises. When I’m training at home, the dogs do not have a collar or leash on. The only time I “formally train” with a collar and leash is preparing for conformation work – hand stacking in particular. Even in conformation work, the dogs learn to free-stack, move forwards and backwards and side to side with tiny hand movements using no leash or collar, just the clicker. The target stick aids in the dog’s moving straight as they will look at the target stick rather than the handler. And, of course, when they are moving beautifully, they hear the clicks!
As there are various books, videos, and information on the Internet devoted to clicker and obedience work, I’ll share some tips that I’ve found useful. Before starting any novice skills, the dog has been introduced to the target stick as described earlier. Using this foundation, when the target stick is held perpendicular against the trainer’s body with the target point up, it encourages the dog to sit closely, squarely, and looking up to the target and ultimately to the trainer’s face. When the dog is moving, the target stick is held at the handler’s side with the target just where the dog’s nose should be. It also helps in cueing the dog in moving into the heel position.
Teaching a dog to heel can be very frustrating and lengthy. “Heelwork” is a complex exercise for the dog and if they do not understand the concept of being on the handler’s left side all the time when moving or stopping (and sitting automatically), the task is extremely difficult. Many years ago, I attended an obedience seminar by a British trainer, Robert Harlow. In teaching heelwork, he started with teaching the dog’s placement at the handler’s left side when the handler was stationary. He didn’t start moving until he was certain the dog understood that “heel” means at the left side. To accomplish this, he did many changes of position so the dog had to readjust.
This past fall, I returned to “formal” training again. I’d played around with clicker training and thought that this would be a good time to re-think my teaching methods with my dogs. I chose my youngest male, Drala, to “experiment” with as I felt that he would be more forgiving of my ineptitudes in figuring out what I was doing with the clicker, target stick, and new thinking. Drala picked up the target game and “fronts” very quickly. He’d already known about “click and treat” from previous fun sessions. I used the target stick to indicate to him that he was to sit at my side. (I sort of twirled it around beside me in the way that a dog would turn and sit.) At first I was reinforcing sits in the vicinity (to put it nicely) but was quickly able to encourage him with voice and hand movements to sit closer and straighter. Soon he’d figured the heel position. Jackpot!!! (Huge reward and praise.) Then that session ended. The next session, he was a little sloppy in his first attempts; then, when a correct sit happened, he was praised/ rewarded with enthusiasm. Then I started making it more difficult for him by standing in front of him (toes to toes) and asking for a “heel.” The learning plateau hit! While he puzzled out what the heck I was asking him to do, he muttered, talked, and tried different places and poses the whole time. It really was a chore not to chuckle! Then, he came and sat in perfect heel position. Another jackpot!!! And the session ended. The third session, his muttering was very little as he quickly went to heel. Time to change criteria. So with him in heel position, I moved forward a couple of steps cueing “heel” as I would in the formal heel exercise and stopped. He moved into position for the click and treat. A few more steps which led to straight heeling and sitting. Jackpot and another session ended. By the fifth session (and I will emphasize these were not done on consecutive days!), Drala was heeling nicely, sitting when I stopped, and starting to watch for change of directions and pace. As he did not wear a collar or lead, my challenge was to keep his busy mind focused on the heeling. Talking to him and using occasional clicks, as he was on the variable schedule of reinforcement, helped. Once this heeling was strengthened at home, our next challenge was to do it elsewhere with distractions.
Seeing how well this method was working with Drala, I started working his litter sister, Sila. Sila is brilliant, loves to learn, and is a challenge to keep one step ahead of her as her comprehension is extremely rapid. Not surprisingly, Sila picked up the heeling exercise very quickly. This has strengthened my thoughts that teaching heeling with a clicker is even more effective than only using halters. Halter work works on the premise of release from undesired pressure (behind the ears and on the bridge of the nose). Halter work is very effective – even more so when it is combined with clicker training.
Over the coldest part of the winter, I didn’t do much formal training with the dogs. Essentially, they had a two month break. With delight, Drala not only remembered his last lesson but his work was sharper. In joining the local kennel club and having a place to work with different dogs and in a different surrounding, the criteria for working Drala (and the others) have changed dramatically. Drala’s first night at the club showed me that I needed to get the halter out again as his focus wasn’t on any work! My three year old was like a puppy in his vocal enthusiasm in meeting other dogs (especially that adorable Papillion puppy that he wanted to play with!) I was able to get his attention for a few minutes and we did some successful behaviors of different things. The halters do aid the dogs in becoming
focused. Clicker training is a wonderful aid, but the dog must be mentally attentive in order for it to be effective. At home, it is much easier to gain and keep attention without all the various distracting sights, sounds, and scents. Distractions certainly give a new meaning to the challenge of keeping a youngster’s attention.
In conclusion
To review, in positive reinforcement of operant conditioning a subject chooses to change its behavior to receive a positive effect (praise, reward, favorite activity). Reinforcement occurs during or immediately after the behavior is done. In positive reinforcement, the subject wants to achieve the reward and changes its behavior until it gets it. With repeated attempts with similar behavior changes yielding rewards, ultimately behavior is changed. This is pure behavior modification using incentives.
The clicker is a teaching tool and like many other aids in training, once the dog has learned the new behavior/task, the clicker can be phased out of use. The newly learned behavior is then maintained on a variable reinforcement schedule with primary reinforcement (e.g. praise or food treat). The trainer can then use the clicker to teach different behavioral tasks. Dogs of all ages, from young puppies to seniors, enjoy learning in this manner.
Many activities can be taught in the same session. As long as the dog is having fun playing “get the handler to click” (and then treat), the interest will be kept up. Like with any “method” of training –or teaching as I prefer to look upon it – keep the sessions short, upbeat, and always end on a happy and successful note – no matter how frustrating
it can be. A basic foundation skill (e.g. a sit) can always be cued in order to end a session on a positive note.
Using the tools of positive operant conditioning with clickers and building upon a solid foundation, a dog can learn a wide variety of skills. Such a dog has been encouraged to think, reason, and remember… all while having fun. Yes, Alaskan Malamutes can become dogs with alphabet soups of accomplishments before and after their names. Best of all, dogs with a variety of skills that they have learned with their trainers are active physically and mentally and are truly Happy Dogs!