Clearing Up Command Confusion
By Gary Wilkes
If you have ever investigated books about dog training, you may have run into some concepts that appear to defy common sense and logic – and you would be absolutely correct. In the world of dog training, there is as much superstition as there is solid knowledge.
For instance, a common practice among dog trainers is to give a command, wait for the behavior to happen and then offer praise that includes the command word. An example would sound like this – the trainer says “Sit” – the dog sits --- the trainer then says “Good Sit.” Trainers who attempt to explain this practice claim that it helps the dog connect the praise to the proper behavior. I hate to burst the bubble, but that is highly unlikely. To understand that type of sentence, a dog would also be able to fathom the following sentence. “Man, that man was selected to man the mandrel, man.” Most humans would have difficulty following that sentence because of the switch between the word “man” as noun, verb and adjective. Expecting a dog to understand that the word “good” is an adjective modifying the verb “sit” and that “sit” is used both as a verb in the imperative mood and a noun, is a bit much. Regardless of the intent of the trainer, the word “sit” as part of the praise doesn’t make any sense because it doesn’t change the consequences of sitting. It’s simply lazy-talk.
If that sounds a little odd, don’t be surprised. It is common for people to assume that words or “commands” cause behaviors to happen and that we can willy-nilly say things and our dog will somehow understand. From that perspective, it makes sense that it is the end result of a behavior that determines how, when and where a behavior will happen again. For instance, if we give a dog a treat every time it sits, the overall behavior of sitting will become more likely – whether we say anything or not. If we decide to stop reinforcing “sits” the behavior will decrease in its likelihood – regardless of what we say. In essence, the words we say are signposts along the road that point to a destination – they can indicate the way to travel, but inevitably it is the destination that “causes” the journey.
Once we realize that consequences control behavior, we can sweep some of the superstition from our training program. For instance, it is also common for people to chant, growl and yell commands at a dog in expectation that the tone will somehow cause the behavior to happen. This is usually an open admission that Fido’s good performance has been insufficiently motivated. The average dog can hear the crinkle of a potato chip bag from three rooms away. Hearing that sound triggers instantaneous and enthusiastic response. If a faint sound can cause a dog to race at full speed to get a potato chip, but must be screamed at in order to “sit,” the problem isn’t with the command – it’s with the payoff.
The way to test our theory of “behavior by consequences” is really very simple. Go get a shallow bowl and some palatable food treats. Touch a treat to your dog’s nose and then move the treat back toward the dog’s forehead. Hold the treat about an inch above the dog’s nose, so the easiest way for the dog to get the treat is from a sitting position. If the dog sits, say “Good” (bite your tongue and don’t add the word “sit”) and offer a treat. Now repeat the sequence about 15 times. On repetition number 16, don’t lift the treat over the dog’s head; merely say “sit” in a soft tone of voice. If the dog sits, say “good” and offer the treat. If the dog doesn’t sit, say “wrong” in a normal tone of voice and try it again. After a 15 minute session, put the treats away and take a break.
Later in the day, stash a couple of treats in your pocket and go about your business. At some point, when your dog is relatively close to you, say the word “sit” in a normal tone of voice. If the dog does, say “good” and offer a treat. If the dog doesn’t sit, say “wrong” and go back to what you were doing. Wait a couple hours and try another training session.
Now we get to the fun part. After creating a pretty consistent “sit” in response to a normal tone of voice, we are going to test the sit-good sit theory. In your next training session, say “bad sit” in a normal tone of voice. If the dog sits, say “good” and give the dog a treat. If the dog doesn’t sit, move the treat over the dog’s nose and lead the dog into a sitting position. Do this about 10 times, so that the term “bad sit” becomes the cue for sitting.
After you have created the “bad sit” cue, it’s time to flip the coin and screw up the phrase “good sit.” On your next session, say “good sit” in a normal tone of voice and see what happens. If the dog sits, say “wrong” and walk away and ignore the dog for 30 seconds. Continue with this until saying “good sit” causes the dog to just stand there and do nothing. Your goal is to have the dog respond to two different patterns. When you say “good sit” the dog doesn’t sit and when you say “bad sit,” the dog sits, every time.
Now, here’s the point of this oddness. After teaching your dog to sit in response to the words “bad sit” you have made something plainly obvious. Your dog doesn’t care a fig about the words you use to describe behaviors as long as the words predictably point to an opportunity for a reward or, if by complying, they avoid something unpleasant. So, the factor most responsible for “causing” a behavior to happen is whatever consequence comes from doing that behavior. To put this more simply, behaviors and patterns that “pay off” are likely to be repeated.
Being a better trainer requires that we understand the simple mechanisms that create and maintain behavior. Within the basic rules of behavior is the concept that, usually, the results of your behavior will affect your future behavior. For instance, if you happen to say “sit” while leaning forward and your dog sits, the chance that you will lean forward the next time you say “sit” increases. Likewise, if you are in the habit of saying “sit, good sit” as the cue to make your dog sit, your dog’s behavior will reinforce your superstition. The secret to cleaning up this misuse of commands is to start listening to what you say and immediately following the word “Sit,” stop. Good stop.