Trainers are often asked to help with stubborn dogs. An owner may describe their dog as obstinate, or willful, or defiant. They say the dog won’t pay attention because he’s too lazy to work. Convinced poor behaviour is a choice, the owner is frustrated and fed up with making excuses for the dog.
But dogs are not stubborn.
Stubbornness requires having a clear understanding of what someone wants and being determined not to comply with their wishes.
When someone says that their dog is stubborn, my experience is that communication has broken down between them and their dog. They believe the dog knows how to do something because they’ve seen the dog do it well before, but the dog won’t perform when the owner wants them to. The owner feels resentful and sometimes embarrassed. They’d dreamed of having a dog that they could take anywhere with them, a dog that would be calm and self-possessed at all times, a dog that would be a true companion, confidant and accomplice in life, but they can’t even be sure the dog will come when called or lie down when told to do so. They’re upset with the dog and they suspect that they’ve somehow failed as owners because they think their dog doesn’t like or respect them enough to behave well.
The truth is that the dog isn’t refusing to comply. She isn’t stubborn. She probably does care a great deal about her owner. She just doesn’t understand what she’s being asked to do.
For example, Rover sits beautifully inside the house, but not when he and his owner go for a walk. The owner thinks that she and Rover are caught up in a power struggle. She loves her dog but she imagines that he’s decided not to listen to her. The reality is that Rover hasn’t learned to sit outside the house or away from home.
One fundamental difference between dogs and human beings is that we think in completely different ways. Humans think abstractly and generalize everything. When we learn to read, we learn to identify letters and words everywhere, all at once. The notion that we can only read in a particular classroom at a specific time of day while facing in the same direction seems ridiculous to us.
Dogs, by contrast, are concrete, literal thinkers. When a puppy learns to sit in the kitchen and the living room, it’s a great accomplishment. The puppy may soon discover that he can sit in other rooms of the house, too, but their human companions rarely recognize that learning to sit in the bedroom and the front entry are also big accomplishments because, for the puppy, they’re separate tasks.
Because we generalize so easily, it may not occur to us that dogs have to learn that “sit” means the same thing in the house, in the car, on the sidewalk, on the lawn, at a friend’s house, on top of a picnic table, on the bed, in the bank, at church, after dark, in the pet supply store, when facing south, east and north-west, even if someone is on a skateboard half a block away and when a squirrel is running along the top of the fence. Sit means the same thing everywhere in the whole world.
I recently discovered that, while Juno is happy to practice anything in the empty bathtub, Penny is anxious about getting into the tub and particularly doesn’t want to go anywhere near the drain. I have no idea why Penny is nervous about bathtubs (she gets bathed by the groomer), but I realized that I haven’t tried to work with her in one until recently. I’m working on desensitizing her to the tub, now, helping her to learn that the bathtub is a good place and that nothing unpleasant will happen to her if she’s in it. It’s really okay to sit, down and stay in the bathtub.
It would be easy for me to imagine that Penny doesn’t trust me or that she doesn’t consider it appropriate to be jumping into and out of an empty tub or even that she has disdain for this silly game I’m trying to get her to play with me. My imagination isn’t helpful when the truth is so much simpler. Penny hasn’t learned to do ordinary things in the bathtub. That’s all. It’s got nothing to do with how she feels about me. She doesn’t have an opinion about whether or not it’s appropriate for her to do these things. She’s never needed to know about working in bathtubs. If I want her to hang around in the tub, or at least not be nervous about it, I have to teach her.
Fortunately, it doesn’t take long to teach a dog to sit in many places. Once the dog has learned that things can be done everywhere in the world, the dog will start to generalize new behaviours more quickly. But if we don’t teach her, the dog won’t know what we want.
Even after doing lots of work, when we think the dog knows how to do something just about anywhere and anytime, there are probably situations which may surprise us. Will your dog sit if you ask her to do it when you’ve turned your back to her? What if you lie down on the floor or stand on a chair?
Occasionally, someone says that their dog won’t work for treats, or that treats don’t make a difference because their dog is stupid, or slow, or uncooperative. The real problem may be that the owner isn’t training at an ideal time, or isn’t using appealing treats. Their rate of treat delivery may be slow, or perhaps they haven’t been reinforcing often enough to make an impression on the dog.
If training time is always after she’s eaten her dinner, the dog may not be interested in more food. She may respond with much greater interest before dinner, however. Many trainers give their dogs half of their breakfast and use the rest for training throughout the morning. Others don’t feed breakfast but use the kibble for training and give the dog whatever is leftover at lunchtime.
While I can do a lot of ordinary training with their kibble, I find that my dogs respond better and faster when I’m offering more valuable rewards like sliced hot dogs, pepperoni, sharp cheddar, blue cheese, tripe, dried sardines or peanut butter. Penny loves frozen peas and frozen green beans. Juno loves whatever Penny is having.
I use high-value treats when I want to make a strong impression, such as beginning something new or working in a different environment. Once the dog has the general idea, I may switch to a less-valuable but still welcome treat. For some things, and especially for teaching the dog to come when called, I always use the best food I have in the house.
Rate of reinforcement
Our culture is moralistic about food and about rewarding accomplishments. We often project those values on our dogs. If someone thinks that giving treats is fundamentally wrong, they may be reluctant to give them and when they do, they may be stingy about it. Their dogs tend to learn more slowly than others, not because they’re unintelligent or don’t like treats, but because they don’t know what they’re being rewarded for. But when dogs are reinforced heavily and rapidly for doing something their handlers really want, they learn much better!
How long does it take you to deliver 20 pea-sized treats to your dog? Two minutes? 30 seconds? 10 seconds? If you deliver 20 tiny treats in 10-15 seconds to reinforce a down-stay or to teach your dog to go and lie down on a mat, you will progress light years faster than if you only give one or two in 30 seconds.
Don’t expect to go from using lots of treats to just a few in one or two days. From 20 treats in 15 seconds you may, over a week or so, gradually shift to delivering 10 treats in 3 minutes, while the dog holds the down-stay. You’ll then be using the treats to teach the dog to build duration into the down-stay.
A dog’s breed also makes a significant difference in training. ((See Stanley Coren, The Intelligence of Dogs (Bantam Books 1994 & 2006). Wikipedia has a good summary of the book and Coren’s methodology, including his survey of 199 North American obedience competition judges which is the basis for his ranking of breeds by intelligence. The article is available at http://en.wikipedia.org/wiki/The_Intelligence_of_Dogs.)) Every dog breed was developed for a specific purpose. Retrievers, for example, have been bred to work closely with humans, ready to perform tasks when directed. As a result, they pay close attention to their handlers. By contrast, terriers and mastiffs have been bred to work at a distance and to rely on their own judgment about what they need to do. The handler may take the terrier out into the field to look for vermin, but the dog finds it, often going underground to get it, and the dog is the one who kills it.
Mastiffs guard property, patrolling the bounds and doing what’s required precisely because the owner isn’t there at the time or, if there, can’t get to an intruder quickly enough. Livestock guarding dogs, like Great Pyrenees, are born in the barn among the sheep. They protect flocks because they consider the sheep to be part of their family, not because a human has told them to do it.
The breeds that are intended to work closely with their handlers generally learn much faster than those who work at a distance. A poodle can learn a new behaviour with just 3 or 4 repetitions most of the time and may be able to perform it at least 95% of the time after that. A cairn terrier may need about 15-20 or more repetitions to learn the behaviour and can be reliable 70% from then on. Beagles usually require 80-100 repetitions to learn something new and may perform that behaviour accurately 25% of the time.
Some breeds are more work to train than others, not because they’re stubborn or defiant, but because they have to learn to attend to human beings in addition to learning what we want them to do. As a result, unless they’re extremely well-motivated, they learn behaviours more slowly and are less likely to perform them consistently.
How quickly a dog learns can reflect the kinds of things the dog is being taught. A beagle may not lie down rapidly on cue, but if the task is scent oriented, the dog will probably learn quickly and reliably because her breed is intended to identify and follow scent. No wonder beagles excel at seeking drugs and explosives in airports! In uncontrolled environments, though, the beagle’s fabulous sense of smell frequently overwhelms everything else to the extent that she simply may not be able to perform many basic skills because she’s so preoccupied by a scent that humans probably don’t notice.
Since repeating learning in many different contexts is so important, an owner can easily become bored or frustrated if their dog needs to have dozens of opportunities to learn to do something in each context. But for those who want to work with breeds that are harder to train, the rewards are tremendous.
And so, if the dog seems to be learning more slowly that we expect, we need to consider:
- Has the dog learned to generalize the behaviour by performing it in many different contexts?
- Is the handler using treats that the dog really loves?
- Are the treats being used correctly in terms of quantity and speed?
- What are this breed’s limitations? Will it take a long time for a dog of this breed to learn the tasks the handler needs? How might the handler increase the dog’s motivation to learn?
November 5, 2014