Beyond Basic Positive Dog Training Techniques
Training shouldn’t stop at “sit!” There is so much more your dog can do!
Look back on past articles in the Whole Dog Journal archives, and you’ll see a wealth of information on basic training and management – two vitally important topics. There comes a time, however, when dog and owner need to move past the basics to more advanced concepts of learning and behavior if they want to develop and enjoy their relationship to the fullest. Let’s explore a few of the concepts that can take you and your dog beyond “Sit happens!” to the hallowed halls of higher canine education.
You will often hear trainers say that dogs don’t “generalize” well. This means that just because Buddy learns to sit beautifully in your living room, he may not necessarily sit when you ask him to at the checkout counter of your favorite pet supply store. He thinks “sit” means “sit in the comfort of my own living room.” You think he’s being stubborn because he “knows” how to sit, when in fact he really only knows how to sit on cue at home; he hasn’t learned to generalize the behavior.
The statement that “dogs don’t generalize well” is actually only a half-truth. Most dog owners can tell stories of “one-trial” learning, where a single experience taught a dog to fear men with beards and hats, or to chase cats that run, or instilled some other high-arousal, strong behavioral response.
For the most part, behaviors that don’t generalize well are those that involve operant conditioning, where the dog acts on the environment. Fido has to learn that he can make good things happen by sitting. Behaviors that involve a strong emotional response such as fear or the chase instinct are quite often learned in a single incident, say, a man with a hat and beard tripped over him and startled him, or a cat jumped out in front of him, hissing and spitting, tantalizingly close, and he gave adrenaline-pumping chase.
In these cases of classical conditioning, where the environment acts on the dog, Fido doesn’t have to learn an emotional response, it just happens. It’s easy for the response to happen the next time Fido sees a man with hat and beard or a cat, even if the man doesn’t trip over him or the cat doesn’t jump up and run.
So how do you help your dog learn to generalize operant behaviors (where the dog acts on the environment)? By doing exactly what your trainer told you to do: practice with your dog in as many different places as possible. In line at the bank. At the dog park. On your walks around the block. In the waiting room at the vet hospital. In the aisles of the pet supply store. In addition, if a dog has truly generalized his “sit” cue, he will sit if you whisper it, yell it, if you’re standing next to him, sitting on a chair, or lying on the floor across the room.
The more behaviors you help him generalize, the easier it becomes for him to generalize each new behavior. Before long, you’ll have a dog who is as well-behaved in public as he is in the comfort of his own home.
Ah, the “D” word – a very bad word in employment or politics, but a very useful one in dog training.
In training, discrimination has nothing to do with skin color. It has to do with teaching your dog to differentiate between one or more relevant stimuli from all the other stimuli in the environment at the time. That means that he sits when you say “sit,” and doesn’t sit when you say “down.” He may learn to bark when he hears your doorbell, and generalize that to all doorbells, including the one on your favorite TV show. You could, if you wanted, teach him to discriminate, and only bark when he hears your doorbell, not any others.
In more complex discrimination exercises, you can teach your dog to distinguish one object from another. In Utility (upper level) obedience competitions, each dog must do a “scent discrimination” exercise. Using his nose, he must find the object that his handler touched amidst a pile of similar objects not touched by the handler, and bring the correct object back to his human.
There was a lovely example of discrimination on Pet Starz recently. A small, elderly Beagle correctly retrieved a half-dozen items from a bag, one at a time, after being cued each time by his owner to get the item by name. The dog was letter-perfect.
You can teach your dog discrimination with objects by asking him to bring you his toys, one by one, as you name them. This skill can be extended to your slippers, portable phone, car keys, etc. This is a vital skill for assistance dogs, and would be a useful thing for the canine companions of any person who has limited mobility.
You can also teach your dog discrimination with locations by teaching him to go to different designated spots. For example, “Go to bed” might mean you want him to lie down in his kennel, while “Go settle” might mean you want him to lie down on his bed in the corner of the dining room.
You can even teach him discrimination with people, by teaching him the names of all your family members, and then asking him to “Find Timmy,” “Find Susie,” “Find Dad,” etc. (See “Targeting Your Way to Discrimination.")
When we say something is salient to a dog, we mean it has noticeable significance to him. Your dog can learn to sit even in the face of distractions because the hot dog you hold in front of his face is very salient. When we associate the hot dog with the verbal cue (“sit!”), the cue itself becomes significant. The salient stimuli in the environment – you, your hot dogs, and the sit cue – are more significant than the distractions. They overshadow the dog barking across the street, the skateboarder whizzing by on the road, the slamming of a car door down the block.
If your dog is too distracted to respond to the sit cue, then the distractions are more salient than you and your hot dogs. You either need to move your training to a less distracting environment, or find a way to make you, your treats, and your cues, more significant to your dog.
This term refers to a phenomenon that occurs when the use of a known cue overrides the dog’s ability to learn a new cue for the same behavior. Keep in mind that, while dogs can only learn one response to a particular cue (“sit” must always mean sit, it can’t sometimes mean lie down), they can learn several cues that all mean the same behavior.
Dubhy, our Scottie, can lie down in response to the “down” cue in English, French, Spanish, German, and two different hand signals. This happened as a result of his role as a demo dog in some of my classes.
I use the “down” exercise to introduce my students to the importance of teaching their dogs to respond to verbal cues without body language assistance. We start by having the handlers lure the down, and as soon as their dogs will lie down easily by following the (treat) lure, we introduce the verbal cue; any new cue you teach must always precede the known cue. I use a demo dog to show them that the dog doesn’t initially understand or respond to the word “down” until we associate it with the luring motion that means “down” to the dog. The motion is salient to the dog; the word is not.
I explain that in order for the dog to hear the word and learn that it also has significance, they must say the word first, then lure the dog down.
If they give the verbal cue at the same time or after they lure, the lure blocks the dog’s ability to learn the new cue.
With enough repetitions of the sequence – verbal cue, followed by lure and click! (or another marker), and treat when the dog performs the behavior – the dog will learn that the verbal cue also has salience, and you will no longer need to lure him down; he will lie down when you give him the verbal cue.
As to Dubhy’s multilingual talents? As soon as he learned a new verbal cue for “down,” I could no longer use that cue to show my students what to do when the dog hadn’t yet learned the word; Dubhy would go down too quickly. I had to keep switching to new verbal cues in order to show them how to avoid blocking when adding a new cue for a known behavior.
Chaining and backchaining
These are two important concepts that come into play when teaching your dog a complex sequence of behaviors. The behaviors are linked together so that each behavior is the signal for the next behavior in the chain. When a musician learns to play a piece by memory, she is chaining; each note or chord draws her forward to the next note or chord in the piece without her having to stop and think about what comes next.
The show ring obedience retrieve is an example of a chained behavior. With her dog sitting at heel, the handler tosses the dumbbell, then gives the cue to “Take it!” Without any further instructions, the dog runs out to the dumbbell, picks it up, returns to his handler, and sits in front of her, still holding the dumbbell until the handler gives the cue to release it and return to heel.
The “retrieve over high jump” is performed in the same manner, except the dog knows to sail over the jump in both directions, going out and coming back, again without further cues from the handler.
With backchaining, you begin by teaching the last behavior in the chain, and then add each step in reverse order, until the dog performs the complete behavior. The theory is that when you teach the last thing first, your dog always moves toward the thing he knows best, so he gains confidence as he learns the new links in the chain.
The song, “Twelve Days of Christmas,” is a classic example of backchaining. You may forget how many “lords a-leaping,” or how many “maids a-milking” but I’d bet you never forget that partridge in the pear tree, and you get faster and more confident in once you get to the five golden rings.
We recently placed a ramp over the three steps from our deck to our backyard so Dusty, our aging Pomeranian, could go up and down more easily. Dusty was afraid of the ramp. I tried luring him up, but he refused to set more than his front two feet on the surface. So we backchained. I set him on the top of the ramp, one body-length from the deck, and lured him up to safety. He did that easily, and after several repetitions I placed him a little farther down the ramp and lured him up to the deck. It took less than 15 minutes to get him confidently running up the ramp. Then we reversed it, and in just a few minutes he was running down the ramp as easily as he was running up.
No, there is no Mack or Postmack, as one of my interns wondered recently. Premack is the scientist (first name David) who in the mid 1960s demonstrated that you can use a more rewarding behavior as the reinforcer for a less rewarding behavior, thereby improving the performance of the lesser behavior.
This principle is also sometimes called “Grandma’s Law,” as in, “You have to eat your vegetables before you can have dessert.” If your dog would rather chase a squirrel than come when you call, you can use the Premack principle to teach him that he will get to chase a squirrel (sometimes) if he comes to you first. Start by applying Premack indoors in a controlled environment, and move outside when he’s doing well.
Leave your dog on a sit-stay and walk across the room. Position a helper with a plate of smelly treats halfway between you, slightly off to one side. The helper should have a bowl to cover the treats with if your dog tries to eat them. Now call your dog. If he stops to investigate the treats, the helper covers the bowl, keeping him from having a taste. Keep calling your dog cheerfully and enthusiastically. When he comes to you, say “Good boy!” and “Go get it!” Race with your dog back to the treats, now uncovered, and let him have some. Then cover the bowl and try again. Eventually – quickly, for some dogs – he will realize that he gets the treats if he comes to you first, and he will fly past the uncovered plate as fast as he can.
There are a couple of drawbacks to using Premack in real life: If the vegetables are too unpalatable, dessert may also lose its appeal; and you can’t control squirrels.
Habituation and learned irrelevance
These two concepts are quite similar. Habituation occurs when a dog learns to ignore an environmental stimulus, such as a startling noise, like the ringing of the telephone, or a disturbing sight, like a realistic statue of a dog. A dog who has never lived indoors may discover all sorts of disturbing stimuli if he’s brought into a household.
Dubhy, the multi-lingual Scottie, was six months old when we found him as a stray and brought him home to join our pack. He had clearly never lived in a house before, and when he saw his reflection in a full-length mirror he spent several minutes, on several occasions, peering behind the door to try to find the other dog. Eventually he habituated to the sight of the elusive Scottie and stopped looking.
Habituation is useful for training because dogs can learn to adapt to stimuli that are initially quite startling and distracting.
However, sometimes the opposite effect occurs – sensitization. Some dogs, rather than habituating to a sound such as the telephone, become more and more reactive each time the stimulus occurs. Thunder phobia is a perfect example of this.
Learned irrelevance, while similar to habituation, applies to a dog who has learned to ignore a cue, rather than becoming accustomed to a startling stimulus. This is not deliberate defiance on the dog’s part, but simply his response to a cue that has failed to have consistent and sufficiently strong significance attached to it. The cue becomes meaningless if it doesn’t have a consequence. It’s not salient.
“Come” is the most common example of this. Many dog owners use this word to call their dogs long before they ever take the time to actually train their dogs to come on cue. By the time they try to teach the dog to come, the dog has already learned that the word has no meaning.
The insidious thing about learned irrelevance is that once it has taken place, it’s very difficult to instill salience to the cue. If your dog has learned that the word “come” has no meaning, it will be easier for you to train him to come with a new cue than to try to make the old one significant. I have heard people use “Close,” “Let’s Go,” and “Here” in place of an irrelevant “Come!”
Principle of parsimony
This scientific principle applies to situations beyond dog behavior and training, but it’s very applicable here too, and one of my favorites. It says, “Unless there is evidence to the contrary, you must account for a phenomenon with the simplest explanation available.” Or, as one of my favorite radio personalities likes to say, “When you hear hoofbeats, think horses, not zebras.”
If your adult dog has a single housetraining accident, the simplest behavioral explanation is that he had to relieve himself. Stress, bladder infections, and tumors are lower down the list, and “spite” doesn’t even merit consideration. The appropriate response is to monitor his water intake and bathroom trips for a week or so to make sure he gets ample potty opportunities. If he continues to have accidents, then a more complex behavioral or medical cause would be suspect.
If your dog often pulls on the leash, the simple behavioral explanation is that he wants to go faster than you do. Dominance and defiance aren’t even in the picture. If he normally walks politely on leash but suddenly starts pulling, simple explanations would be that something frightened him and he’s trying to get away from it, or something very enticing is in front of him and he’s trying to get to it. Again, dominance and other complex motives are unlikely.
As you can see, training can be a little more complicated than the basic “sit,” click!, and treat. The more you learn about the workings of your dog’s brain, the better you’ll understand how and why he does what he does, and the better prepared you are to respond appropriately. It can only enhance your already wonderful relationship with your best friend.
-Pat Miller, WDJ’s Training Editor, is a Certified Pet Dog Trainer, and past president of the Board of Directors of the Association of Pet Dog Trainers. She is also the author of The Power of Positive Dog Training and Positive Perspectives: Love Your Dog, Train Your Dog. See “Resources” for contact and purchasing information.