BarkingDogs.net

Go to the index for this article

Go back to page five of this article

Page Six of a twelve-page article:
A Detailed Examination of the Process of Bark Training a Dog

Training Your Dog Not To Bark

A procedure that is worked right will work right, because exact procedures get exact results. On the other hand, sloppy procedures, at best, get some results, sometimes.

I used to teach a psychology course in which I introduced the students to learning procedures by having them work with their dogs at home. I emphasized that the way to train a dog not to bark was to do four things after each and every violation:

1. rush immediately to where the dog is (do not call him and have him come to you)

2. say "no," and

3. follow the "no" an instant later with a punishing smack on the nose.

4. then, make an immediate retreat back into the house.

One of the students gave it a try and reported back that the barking increased after she started training, and that the more she worked with the dog, the worse the problem became. She insisted that she was doing everything just as I instructed, but when she described her interactions with the dog, it became apparent that the intervention she was using was not the one I urged her to try. Her intervention was as follows. Every time her dog barked she would:

1. rush immediately to where the dog was

2. say "no," and

3. follow the no an instant later with a light tap on the dog's hindquarters.

4. then she played with the dog for a minute before going back into the house.

She did well on steps one and two, but what she did on steps three and four was very different from the intervention prescribed. Only exact procedures are guaranteed to work. Inexact procedures, like the one she used, produce, at best, some results, sometimes. In this case, her perversion of the intervention proved counter-productive.

It's true that the smack on the nose is supposed to be light enough so as not to be particularly painful, but it still needs to be delivered in such a way that it is very unpleasant. My student barely tapped the dog on his hindquarters, which was not at all aversive to the animal, so it had no effect. Likewise, saying "no" to the dog had no effect on that particular dog during that phase of his training because "no" only works as an aversive after it has long been paired with and, thereby many times associated with, something unpleasant, like a punishing smack on the nose. All of those factors came together to make her intervention ineffective. What made it counter-productive was that the dog's owner socialized with, and dispensed affection to the dog before returning to the house, which meant, in effect, that every time the dog barked, he received a reward. When a particular response (like barking) is consistently followed by a reward, the response is expected to increase, which is just what it did.

The problem with behavioral procedures is that the effectiveness of any given procedure is limited by the degree of precision with which it is executed. The point being that if you want this intervention to work, you need to follow the formula exactly. Changing seemingly small details of the intervention can make for very large differences in your results.

Eliminate the Reinforcers That Support the Barking Behavior

In your effort to bark train your dog, your first step needs to be to eliminate the rewards (the reinforcers) the dog receives for barking. To determine what is reinforcing your dog's barking, check to see what happens immediately after he barks. If his barking is followed by some sort of attention or by a turn of events he finds interesting, then that is probably the reinforcer.

One of the basic rules of behavior, one that is almost always true, is that if someone is making a response, then there must be something that is reinforcing that response. It's like a magic act where the magician makes his assistant levitate in mid air while he passes various hoops and rings around her body to show there is nothing holding her up. Even though your eyes tell you she is levitating, you still know that something is supporting her because, otherwise, she would fall. It's the same sort of thing with any given response. If the response exists then there will almost always be a reinforcer present to support it and keep it going. That's sometimes not true with barking dogs, however.

A stir-crazy dog may bark in the absence of external reinforcement simply because barking at nothing in particular is slightly more entertaining than standing around doing nothing at all, which is his only alternative. Also, a dog vocalizing in a threatening way may bark simply because, by virtue of his genetics, he enjoys hurling threats at passing strangers.

Stretching the Ratio

Dogs that seek social interaction by beckoning to people in the distance may also bark in the absence of reinforcement. That's due to a phenomenon called stretching the ratio, which I alluded to earlier. The term "ratio" refers to the ratio of response to reinforcement. In other words, how many responses (barks) does the subject (the dog) have to make before he receives a reinforcer (some human attention). Very often, when people first get a dog they will frequently yell at him to shut up when he barks. With a lonely, desperate dog, isolated by himself in the backyard, that may be all that is required to reinforce his barking behavior. He may learn that by barking he can draw a verbal response from you, which makes him feel less isolated and abandoned. After a while you become less inclined to answer his bark with your own verbal response. So that, as time goes by, you respond to his barking less and less often. For some reason, gradually withdrawing reinforcement in that fashion causes the response to continue even after the reinforcement has been faded out to next to nothing. So, by stretching the ratio of response to reinforcement. you may eventually reach the point where the dog continues to beckon endlessly even though people seldom, if ever, respond in any way. So it is possible that, at this point, your dog may be barking in the absence of any reinforcers that are within your control.

Still, all types of barkers are likely to receive at least some reinforcement from passing pedestrians who reinforce the barking by looking at the dog or reacting to him in some other way. In that instance you can identify the reinforcers but, outside of blocking the dogs view, you can't do much to eliminate them, not when there is a steady stream of people passing by within view of the dog.

In any case, if you can identify any reinforcers that may be driving the barking response, begin by removing them. It is much easier to punish a response out of existence if you first remove the supporting reinforcers.

Reinforce Incompatible Responses

A dog can't both bark and not bark at the same time, which means that you can reduce the amount of barking by reinforcing instances of silence or, in behavioral terms, you can reduce the problem behavior by reinforcing an incompatible response.

You can eliminate a barking problem through punishment alone. It's not necessary to reward the dog for not barking by spending time with him when he's quiet. But it will be easier to stop the barking, and it will be less upsetting for the dog if the alternative to barking is waiting a little while for you to come out and join him, than it will be to stop the barking if the dog's alternative is to sit forever by himself.

It may have occurred to you that you can't reinforce a dog for being quiet if he never stops barking. When you are dealing with a dog that barks virtually without letup, you have to punish the continuous barking until you reach the point that the dog is sometimes silent. Then, you can begin to reinforce him for quiet behavior.

Go forward to page seven of this article

Go to the index for this article

This page on bark training is part of Section One:
the Your Dog section of barkingdogs.net