Chapter 2. Every Dog Learns

The Association Model for learning. The Association Principle can explain virtually all learning. Learning happens when new actions are associated with old ones. Actions are events that trainers arrange in the dog's training setting. You will teach your dog the meanings of some useful actions that the dog can sense going on in his surroundings. To be an action, it must affect the dog in some way, i.e., make an impression, be perceived, rouse the dog to activity, and/or stimulate an emotion. An action can be a voiced command that is heard by the dog, a touch that is felt, a human scent that is smelled on a track…

The Association Principle is defined in a way that makes it useful as a training procedure. For our purpose, the principle is expressed as the close pairing of two distinct actions.

For association learning to work, two requirements must be met: 1) the two programmed actions must appear in the correct order; (2) the interval between the start of the first action and the start of the second must be quite short, yet not overlap.

   A0 is followed by A1

Figure 2-1. The Association Model.

In this model, A0 and A1 represent actions. A0 is an action that either has no meaning to your dog or has a meaning that you want to change; A1 is an action that means something to the dog, and which may even control his behavior in some way. Simply stated, the association principle says that if an action, A0, appears consistently before another action, A1, then the first appearing action (A0) will take on the meaning of the one that follows (A1).

Consider this example: Suppose you suddenly snap your fingers (A0). Your dog may show a mild reaction to the sound when you first do it. But when you do it again, the dog might not react at all. From this, you know that your dog hears the sound, but apparently, it has no particular meaning for him. Do it a third time, but now after you snap your fingers, get up quickly, grab the leash, attach it to the dog, and proceed outdoors. You have just performed an association procedure: the finger snap is A0, and going-out-of-doors (a pleasurable action) is A1. When you do the pairing one or more times, your dog should be leaping with excitement whenever he hears the finger-snap.

If the dog is to learn the meaning of the first-appearing action, A0, he must at least be aware of its presence. When you see that your dog pays no attention to one of your actions during an early training period, you must emphasize the event in a way that it's noticed, i.e., by giving a louder voice command, exaggerate a signal, etc.…

Speed of learning an action is most favorable when the time interval between the start of action A0, and the start of action A1, is in the range of several milliseconds to about one-half second. You will show good training technique when you start the second action about one-half second after the start of the first. Do not begin both actions at the same time—that is, don't overlap the beginning of both actions. Also, do not do them in the reverse order, otherwise the desired learning will be delayed or not occur at all.

If you don't have the knack to "time" the actions too well, you may increase the time interval between actions to one second. Training will just take a bit longer. Better yet, do Team Training with your dog (see the section, Team Training, in Chapter 1). Have another Team Player deliver one of the actions in any pairing.

Again, by Association Learning, you have the ability to change the meaning of an action, as perceived by the dog, by properly sequencing and timing it with an action that has the desired effect. The procedure is used extensively in dog training.

Accidental Associations

We know that associations between actions can be made by chance. Consider this coincidence: With outstretched arms, you call the dog to you. When he comes, and a split second after you touch him in a petting gesture, a frightening noise sounds. You can imagine what the dog will do when you reach out to pet him again the next time.

Arranged Associations.

Arranged associations are what you will be doing in this training course, where it's assumed you have control over the training setting and training actions, and, consequently, over what the dog learns. In this instance, associations between actions are made by design, and not by accident.

When you arrange an association between two actions, the first appearing action can be said to become a predictor of the second, as was cited above in the finger-snapping instance. One of the first associations that you will make in this course will be between the voiced GOOD and a food delivery—in which case, GOOD will become a predictor that a food tidbit will follow.

Learning New Commands and Signals.

Your dog will learn new commands (and signals) that will be paired with ones that the dog already knows—the latter may already be controlling a dog's response that is of interest to you. For example, if you want the dog to learn the meaning of the voiced command DOWN, and the dog already knows to go down to an arm-and-hand signal, then you would pair the new "command" with the familiar "signal" in this order: the new command DOWN comes first, followed closely by the arm-and-hand signal. When paired this way over a number of associations, the command DOWN will come to control the down response when given alone.

Altering an Emotion

With the Association Procedure, you also have the ability to change the emotional character of a first-appearing action to that of a subsequently appearing, more desirable, one. That is, if the first-appearing action of a pair is fear-eliciting (A0), and the second action is pleasure-eliciting (A1), then the consistent and close pairing of the two actions in that order will result in the first-appearing one, A0, also eliciting pleasure.

As an example, hunting dogs that have a fear of gunfire can be helped to overcome that fear: Something pleasant can be made to happen to the dog whenever a gun is fired nearby. For instance, a tasty morsel could appear immediately following gunfire. The dog would soon look forward, with anticipation to hearing a gun sound.

If you reverse the order of the two actions, then the pleasure-eliciting action will become fear-eliciting. That is, suppose the trainer fired the gun immediately after giving the dog a tasty tidbit. In only a few such associations, the dog would fear to pick up the tasty morsel. So, watch the way you order the actions (provided you have control over them in your training setting). It makes a difference.

top ::: home