Most dog owners think they know how to train a dog in the same way that most parents think they know how to raise a child. The truth is, some Training does come naturally. When children or dogs do something that results in a caretaker becoming uncooperative or any other consequence that is uncomfortable, the child/dog does that thing less often in the future. After many trials, some unwanted behavior disappears completely. However, the behavior has to have no value of its own or it will be repeated when caretakers are not supervising. This leads to the consequence being associated with Supervision instead of the behavior; a considerable weakness of osmotic training.
The same can happen with behaviors that result in us caretakers suddenly paying attention to our students, even offering tangible rewards. The behaviors that result in this positive outcome are likely to be repeated. If we are very good trainers, we will set up situations in which our trainees are likely to do good stuff so we can reward it, attach a name to it and ask for it whenever we choose. But if we hang around supervising and waiting for stuff to happen, the full limitation of training by proximity is apparent.
Many of the behaviors we want our dogs (or kids) to engage in, don’t happen by accident. Even if they do, they don’t happen often enough to be connected with their rewards. So we must elicit behaviors we want in order to teach dogs anything useful.
One method of eliciting a behavior is to pressure a dog into a certain action and take the opportunity to reward that behavior. Some trainers would argue that the release of the pressure is reward enough. This is called negative reinforcement–when a discomfort is removed if a dog gets it right. But this technique has some of the same fallout associated with punishment. If used when a dog is just learning a new behavior the discomfort of pressure is associated with that behavior, its cue or training sessions in general.
A much better way to teach a new behavior or part of a chain of behaviors that will be useful, is to create a situation in which the behavior you would like is impossible to avoid. Trainers call this “error proof” learning. For example, if you want your dog to learn to walk at your side, you set up a situation in which walking at your side is going to happen. And then you reward it often enough that it is repeated with purpose. Then you call it “heel” and ask for it by name.
An exercise called Connect the Dots is popular for teaching heeling. Tack down some paper plates on the ground at 6 foot intervals. Walk your dog on a leash past the plates. Stop at each plate while handing your dog a treat. Repeat. Before you know it, the dog won’t stray from your side because treats keep happening. As you remove every odd plate until your dog watches you while walk when there are no plates, introduce the cue. Ask for the behavior around obstacles then in new places, then with a distraction nearby.
Another popular method to teach the same thing is using a marker for the behavior of being next to your knee. Whenever your dog is accidentally near your knee, you’ll say “Yes!” or click a cricket toy until he starts getting into “near knee” position on purpose and looking for the marker or treat. Once he does this with purpose, use the cue as you move randomly about; backward, to the side, one step to the right, until no matter what you do your dog chooses to be near your knee. Gradually increase the number of steps between marks and rewards until your dog is heeling.
Either way you create the opportunity to get it right. Once your dog is at about 80% correct responses to the cue, add some distractions. Here is where pressure and release comes in.
When you ask your 80% correct dog to heel in trying circumstances and he does not comply, adding some kind of pressure that is immediately relieved by assuming the heel position is what is meant by pressure and release training. For example, a forging dog may be pressured by ceasing to move forward until he allows for slack in the leash. Some trainers say if error proof set ups are carefully designed, no pressure is necessary to ensure a dog knows the cue and will not fail to comply even in distracting circumstances at a distance. Some trainers, especially those involved in police or search dog work, are reluctant to trust a dog to comply in trying circumstances without knowing for sure the dog understands there is more to lose than the treat for ignoring a cue. And this is where it all goes south.
The arguments about this are common, can be brutal and personal, and do very little to resolve the issue of what is right or wrong. Understand the issue so that apples and oranges and straw men or name calling don’t get in the way of answers. Reading your dog’s body language and measuring your success as a team is the way to determine how you will train. But you’ll have to at least deal with these issues to avoid the pitfalls of training by osmosis.