Recently I watched a very skilled trainer work at shaping a dog to perform a simple task. The trainer’s timing and mechanical skills were simply astounding, and it was a real treat to watch her ability to split a complex behavior into simple, easily achievable steps. The dog learned quickly and was soon performing the target behavior.
In spite of all of this, I had a nagging sense of unease as I watched the team work. Finally, it hit me. There was no connection between the dog and person. Each time the dog looked into the woman’s face (which he did whenever he was confused and each time he earned a reward), he was met with blankness. When the dog figured out what the woman wanted, she clicked and tossed a treat, but her facial expression never changed. Watching the interaction between the two was like watching a lab animal being trained by a Skinner box or a robot. The techniques were sound, but the training was sterile.
This style of training is not uncommon, and it’s rooted in a very sound and important training principle. However, some trainers seem to take it to the extreme, to the detriment of their relationship with their dog. Let’s bust the myth of emotionless training once and for all.
Many novice trainers chatter at their dog, and there’s a good reason this is discouraged. Learning is difficult, and it’s even more difficult when someone’s constantly talking at you. Imagine trying to solve a complex math equation in your head. Now imagine that as you’re trying to figure out that equation I start encouraging you.
“That’s the way!” I tell you, “Keep working! Coooooncentraaate… good! Now carry that three… hold it… hold it… almost there… way to go! You’re so clever! Keep concentrating, you’re doing great!”
Does my encouragement help you solve the math problem? Of course not! In fact, the added distraction of my voice probably makes things even more difficult for you (unless, of course, you’re able to tune me out).
This is the sort of thing that dogs with novice trainers are subjected to all the time. Chattering at your dog as he’s trying to figure out what to do hinders his learning, and likely frustrates him to no end. Problem solving requires focus and concentration, and the more you talk, the harder it’s going to be for him to get it. Remember, dogs don’t speak English. Explaining what you want him to do (“Fido, sit. Sit down for mommy now. Siiit. Sit! Sit! Put your butt down now, mister…”) isn’t going to make the task any more clear.
When we’re training our dogs, it’s important that we keep the static out. Set your learner up for success by reducing or eliminating as many distractions as possible in the beginning stages, including your [very human] need to verbally explain things. In clicker training, especially when shaping a dog, it’s important for the trainer to keep her body still and quiet.
That said, it makes me incredibly sad to watch a dog make a huge leap in learning only to be rewarded with nothing more than a sterile click and treat. Let’s go back to our previous example of the difficult math equation. After you solve the math equation in your head, how would you like me to respond? There’s a big difference between how you’re likely to feel if I just say, “good job” in a flat monotone and if I tell you, “that was VERY well done!” with genuine warmth in my voice. Which reaction is more likely to motivate you to try solving a second math equation for me?
If your dog has done a good job for you, let him know! Make it very clear how proud you are of him. Once you’ve marked his good decision with the clicker, you’ve started the reward sequence. My preference is to make the reward equal to the amount of work the dog had to put into the behavior. This means that I will provide different feedback or rewards for different responses. If we’re shaping a behavior that requires multiple quick repetitions, I will usually just click, smile, and toss the dog a treat. If he makes a leap in learning, I will warmly praise him after I’ve clicked in addition to giving him a treat, telling him how brilliant and wonderful he is. If he’s done something especially hard, I may break out of training for a minute to get down on the ground and play with him or pet him, depending on what he likes.
Many people tell me they want to get away from using so much food in training, and this is one of the easiest ways to do so. Here’s the thing: if you associate your genuine praise with the click and food treat over and over, your dog will start to respond positively to your praise alone. If you don’t ever praise your dog or make a fuss over him in training, though, this transfer of value can’t happen. Click, begin to praise, then give your dog a cookie (in that order), and soon you’ll start to see him responding to the praise as eagerly and happily as he responds to the cookie.
Clicker training is based on very sound scientific principles, and most of the time that works in our favor. However, it is absolutely possible to become too wedded to the sterility of the scientific process, and this is where things tend to break down. We are not robots, and our dogs have a much deeper and more meaningful relationship with us than a laboratory rat or pigeon will ever have with their food hopper. Humans and dogs are both social species, and we can both find joy in living with and working alongside one another.
So go ahead, praise your dog when he does a good job. In fact, throw a party for him! Training should be fun for both of you.