Is There Something Special About Clicker Training for Dogs?
Do you know that clicker training for dogs does not require a clicker?
Posted Mar 02, 2016
I was watching an intermediate level dog obedience training class where the dogs were being trained to respond to hand signals. Floating over the usual class related sounds produced by people's voices and dogs was an array of intermittent clicks. These sounds were being produced by a device held in the hand of a woman who was working with a Malinois. The device was a version of a common toy which first became popular near the beginning of the 20th century that was generically called a "Tin Cricket". These toys contained a piece of spring steel that you could press to make a clicking sound similar to that made by a cricket.
The instructor came over to the woman and said something on the order of "You really don't need to use the clicker in class. Your voice will work fine."
The woman protested saying, "I was told that the best way to train my dog was to use clicker training for my dog and she simply won't learn using any other method. Without the clicker teaching her anything is hopeless."
In recent years there has been a lot of attention given to "clicker training," which is viewed by some people as a more efficient way of training your dog. Unfortunately there is a lot of misunderstanding associated with this method of teaching dogs. Before I try to clear up some of the confusion associated with clicker training it is important to understand that the basic principle behind dog training is really quite simple. That does not mean that the successful application of this principle will be easy. Remember that the basic principle behind playing the piano is simple, all you have to do is to press the keys corresponding to the notes that you want to hear. However years of practice may be required before you can do it right.
The basic principle behind teaching your dog any new task is that any behavior that is rewarded will be strengthened and the likelihood that it will appear will increase, while any behavior that is not rewarded will be weakened and the likelihood that it will appear will decrease. That's it! To train a dog, it is not necessary to understand the underlying neurological or chemical events, or know which brain centers and pathways are involved. The “tricky” part, however, is often being able to get the animal to perform the behavior in the first place so that we can reward it. It is also necessary to get that reward to him at the right time, so that the appropriate behaviors will be strengthened. This process sometimes requires a fair bit of time and effort.
As an example, suppose that we are trying to teach our dog to come on command. What we need to do is to reward him when he is moving toward us after we have given the command "come". However if I rush out to give him a bit of kibble as a reward he will obviously stop moving forward in order to get the treat into his mouth. Notice that this means that the reward has actually interfered with, or stopped the behavior that we were interested in having the dog perform. What we were looking for was a sequence of events where the dog is rewarded while coming toward us, however what we really got was that the dog was rewarded when he stopped moving in our direction. Because the basic principle of dog training is that the behavior that occurred immediately before the reward is what is strengthened, we would actually be teaching the dog to take a few steps toward us and then stop. What we need then is a reward that doesn't stop the behavior, and can be precisely delivered while the behavior we want is occurring. Furthermore, we want a reward that we can deliver from a distance.
Clicker training is based on creating a special reward of this nature. Technically it involves the classical conditioning of emotions. After all, it is not the actual piece of food that we use as a reward that is important, but rather the fact that that piece of food produces a pleasant emotional response in the dog when he receives it. Here, the idea is to produce a signal which makes the dog feel good in the same way that receiving a treat will, but does not interfere with his ongoing behavior.
To create this new reward you first have to decide what kind of signal you want to use. Over the past few years the sound made by that hand held clicker has become popular, but the signal can actually be anything, a whistle, light, word or any specific action. I prefer using my voice since that leaves both of my hands free for other things, and I use the word "yes" said in an enthusiastic tone, however I know a dog trainer who uses the word "click" as a signal. The important thing to note is that there is nothing special about the sound made by a metallic clicker, since any signal which is easily seen or heard will do.
Once the signal is chosen you have to "charge it" with positive emotional value using classical conditioning. The easiest way to do this is to use food rewards. Suppose we have chosen to use that mechanical clicker. To turn this sound into a reward all we need do is to repeat the sequence of a click followed by a rewarding treat a number of times. After we repeat click-treat, click-treat, click-treat, enough times, the clicking sound will take on many of the positive properties of the actual food reward.
Using this process we have now created a learned or "secondary reward." It may help you understand better if you recognize that most of the rewards used to control human behaviors are secondary rewards. Basic or primary rewards are biological, such as food and drink and sex. Secondary or conditioned rewards include money, grades in schools, praise, promotions, medals, titles, and so forth. The rewarding qualities of all of these have been learned through association with primary rewards, although the chain may reach back into the distant past. Thus the rewarding value of praise may trace its origins back to being fed by your mother who accompanies the placing of food in your mouth with words like "that's a good boy" or "what a pretty girl." The importance of creating a signal, such as a click or the word "yes", which has positive emotional qualities, is that it can be used to reward the dog without interfering with his behavior, and later when convenient, it can be followed by an actual food reward to keep it emotionally "charged."
What should be clear is that clicker training is not actually a teaching method, but rather a technique for creating a special reward that can be used to assist us when we are training a dog. The hard part really is getting the dog to produce the behaviors that we are interested in so that we can reward him. Sometimes that can be as easy as simply catching the dog spontaneously producing the behaviors we want and then rewarding him — a training method called "behavior capture" (click here for more on that). For more complex behaviors one may have to use other methods such as "luring," "physical prompting" and "shaping". However, one must remember that "clicker training" is not really a training method but rather a method of delivering a reward in association with other training procedures.
I sometimes think that people feel more comfortable using a mechanical device such as a clicker, because the idea that a simple word of praise applied with appropriate timing will suffice to train a dog appears to be "too simple" to actually work. It is much like the "magic feather" which was given to the elephant Dumbo to convince him that he could fly. It wasn't necessary, but it made him feel good, and confident enough to continue the process.
Stanley Coren is the author of many books including: The Wisdom of Dogs; Do Dogs Dream? Born to Bark; The Modern Dog; Why Do Dogs Have Wet Noses? The Pawprints of History; How Dogs Think; How To Speak Dog; Why We Love the Dogs We Do; What Do Dogs Know? The Intelligence of Dogs; Why Does My Dog Act That Way? Understanding Dogs for Dummies; Sleep Thieves; The Left-hander Syndrome
Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission