The Rat is Always Right. Now, What Was That Cue Again?

Funnily enough, dogs don’t speak English.

This is probably not news to most people, but there are a lot of things about language that dogs don’t understand.  The part of the human brain that processes language allows us to infer multiple meanings from a single word, but dogs’ brains simply aren’t hardwired for language conceptualization.  Not only are our words a foreign language to them, but they don’t have a verbal decoder built into their prefrontal cortex like we do. Think how confusing it must be for a dog to be expected to know that “down” means “lie on the floor,” except when it means “get off the couch,” or “take your paws off my leg,” or any other situation where we might say the word reflexively.

This is one reason consistency is important in training. When we teach our dogs cues (of any type), it’s important to make sure that each cue has only one meaning.

(And if you’ve read my previous posts, you’ll know that this is the part where I give a personal example of what happens when you fail at this. I’m usually a competent trainer — really! — but these facepalm moments make for more entertaining blog posts than “I did something right today!”)

First, the backstory: It took me years to teach Valenzia to retrieve. Not only did she have zero natural inclination to pick up toys or other objects, but she had been (before I adopted her) physically punished for showing interest in anything a human was holding. Normally it’s a comparatively fast process to back-chain a retrieve, starting with rewarding the dog for holding the object and then working back through the previous steps of finding, picking up and bringing, but for months I just couldn’t get my dog to put anything in her mouth.  Training a retrieve was a long process of shaping nose-touches, then nose-touches with an open mouth, then mouth contact, and eventually picking something up. It took a long time, but it also meant that by the time she learned the full retrieve, all those clicks had built an incredibly strong reinforcement history for picking up objects.

Since Valenzia and I were both thrilled to have a fun new behavior, I started asking her to retrieve all kinds of random things around the house. My informal cue was “get it,” which I used loosely to mean “pick up the object on the floor that’s nearest to you.” She thought this was a great game, and would bring me anything from her Kong toys (which I would stuff with food as a reward) to pieces of trash (which could be traded for a treat — also a good practice of object exchanges to prevent resource guarding).

Tonight Valenzia apparently discovered some lost bit of treat under the foot my bed, because she suddenly started rooting around and shoving storage boxes out of the way with her nose. Curious about what was there, I bent over and conversationally asked, “What are you looking for?” (Yes, I talk to my dogs.)

That distant memory of punishment still affects her confidence, so she immediately jumped back, afraid she might be in trouble.  I wanted to communicate that she was allowed to keep looking, so without thinking I said, “It’s okay. Get it!”

Valenzia blinked at me for a second, then began frantically searching for something to retrieve. She poked the storage boxes, but they were too big. There was a basket on the floor, but it was full of things. After spinning around to scan the area a couple of times, she seized a sock from the nearby laundry pile and brought it to me with a quizzical look that seemed to say, “I know I’m normally not supposed to bite clothing, but is this what you want?”

So, naturally, I grabbed my phone and recorded the moment for posterity.

“Is this what you want?”

And then we played some more retrieving games for dog biscuit pieces, because after all, she’d done just what I asked.

* The title of this post, “The rat is always right,” is a famous B. F. Skinner mantra.  In essence, it means that a training subject only does exactly what you have set it up to do. If an animal responds differently than what you expect, you probably haven’t trained what you thought you were training.