I’ve seen a few remarks from people who do not practice clicker training that reflect a fundamental misunderstanding of what it actually is. So I thought I’d address that today. Here are two representative remarks:
- I bought a clicker, but my dog doesn’t do anything when I click it.
- Clicker training creates robot dogs.
I suspect that the origin of these statements can be found in two places:
- Comic strip-inspired movies where villains transform people into robotic killing machines that are activated by a sound or keyword after brainwashing.
- An incomplete and erroneous understanding of the link between clicker training and Pavlov’s research.
Contrary to what the statements above imply, the clicker is not a remote control. It’s a means of communication. You can actually substitute the word “yes” for the click. However, the problem with the human voice is that it is variable and often reflects our emotional states, including stress and worry. Plus, we use words like yes in different settings. The click is invariable and is used only to tell the dog, “What you just did is something I want more of.”
Most people know that Pavlov conditioned dogs to drool on cue. Their mistake is to think that the clicker is a cue. It’s not. AFTER the dog performs a desired behavior, you click and then reward the dog. Just like in Pavlov, the noise predicts the food. The difference is that Pavlov’s dogs just had to sit around, but a dog in training has to offer some sort of behavior.
Using the clicker has another advantage: our mammalian brains are wired so that anticipation of good things actually feels better than getting good things. Anticipating something pleasant stimulates our brains’ pleasure centers more than the pleasant thing itself does. That makes clicker training more powerful than simply giving treats. (And it explains why events never quite live up to our excitement leading up them.)