Basic dog training tends to focus on behaviors with names, or what are often referred to as commands: sit, down, come, heel, stay, etc. But in my experience, about 80 percent of what most people expect a well-behaved dog to do is not behave, or at least not behave in ways that come naturally to dogs. Do nothing when a stranger stomps up the front steps and pushes papers through the mail slot. Do nothing when another dog appears across the street. Do nothing while the rest of the family gathers around a table full of food, delicious food.
Our inclination when we want a dog to do nothing can be to wait for and then punish the thing he's doing instead. It’s hard to wrap your mind around how you can teach a dog to do nothing using positive reinforcement—after all, by definition, there’s nothing to reinforce, right?
Griffon offered his “do nothing” behavior on the welcome mat at the motel on a recent road trip. “The mat is everywhere!” his dad reported.
Sometimes our language seems to get in the way of our eyesight. “Doing nothing” is never doing nothing. Whatever your dog is doing when he’s doing nothing—standing around, lying belly up and staring into space, glancing at something without bothering to investigate it—that’s all behavior too. And behavior doesn’t have to have a formal name attached to it for us to identify it, reinforce it, and thereby make it ever more likely in the situations where it has been reinforced.
Dogs come knowing how to do pretty much any behavior you might want. They just don’t know that you prefer sit, down, and that cute head cock over barking, jumping, and playing tug with your scarf. But when you begin systematically selecting behaviors you like, by immediately following them with things the dog likes, the dog learns to resort to them more quickly and more frequently—often in place of other, less desirable choices.
What's more, the context in which offering that behavior was reinforced will begin to signal to the dog that this behavior is a good choice in that context. In the video to the left, three-month-old Milo, who had been reinforced periodically for settling on a bathmat over the past couple of weeks, was beginning to learn that settling on that mat is a better choice than jumping in the dishwasher to lick plates or jumping up to grab towels as his mom tries to fold them. She's not waiting for him to jump and then kneeing him in the chest or pushing him down; she's proactively teaching him what TO do when she does dishes or laundry. And though it looks like "nothing," it's decidedly not. Likewise, in the video below, young Frida is learning in achievable increments that settling on a towel is better than playing with shoelaces.
Below, young Ralphie is learning the same lesson, with not only shoelaces and laundry folding but also cleaning up pee (paper towel moving erratically on the floor, whee!) and twirling in a floaty skirt. Meanwhile, off camera, his sister Muppet gets reinforcement on her own bed, where she can't horn in on Ralphie's training.
And here Sunny learns a new response to vacuuming:
To teach a dog to do a behavior more often, you can use a technique called "capturing." Applied animal behaviorist Kathy Sdao has described this as See, Mark, and Reward Training (or SMART). I like this, because it emphasizes that you must first be able to notice the behavior that you want to increase. And I love the SMART x 50 game she suggests in her book, Plenty in Life Is Free, in which you set aside 50 pieces of kibble or treats per day, watch for behaviors you like, mark them (with a word or a click), and deliver one of the treats. (The advantage of marking, rather than just delivering a treat, is that it can bridge the gap between when the behavior occurs and when you can get the treat out and give it to the dog.)
Do this with a new dog or puppy for the first few weeks, when many newly adopted dogs are still in the relatively subdued “honeymoon period,” and, without ever barking a single "command," you’ll have applied the power of more than 1,000 reinforcers to significantly increase the likelihood of your preferred behaviors happening, in the circumstances where you would most like them to happen.
Cues like words and formal hand signals can be attached to these behaviors, if you like, once the dog is consciously and confidently repeating them in a predictable way. For sit, e.g., you could simply start to say “sit” right before you can tell the dog is about to do it again anyway. Just like other, environmental context cues, the word can become associated with a high probability that reinforcement is available for that behavior, and can then be used to ask for it.
Sitting is a great alternative behavior to jumping on guests or running out the front door. But so is standing—a behavior we rarely notice and almost never name, because to many of us it looks like "doing nothing." And while most dogs will need quite a bit of training to learn to sit in exciting situations, standing is the behavior that is usually already happening right before a dog jumps or runs. Using capturing—with treats or other reinforcers the dog may be seeking, such as petting—it's fairly easy to teach your dog to simply keep his front feet planted as people approach or as you come in the door. Which would you prefer upon coming home—having to tell your dog what to do or having him figure it out from the context?
Here, Ginger is marked for simply looking at laundry is folded to start with, but starts to offer other behaviors like sitting and lying down on her own:
Likewise, you can capture silence as mail comes through the slot, a cursory glance at another dog, or lingering at the threshold of the kitchen or the baby’s room. A while ago, when I used to routinely leave the house before my husband, I set my sights on one of my own dog’s annoying behaviors: barking at me when, after gathering up my belongings, I walked over to the bed and leaned in to kiss my husband good-bye. I began by leaning in only partway and clicking before Pigeon had a chance to bark, then leaning back out (what I hypothesized was the reinforcer for her barking) and delivering a treat. After a couple days of this, when she was anticipating reinforcement for just watching me lean in, I advanced stepwise through the whole cycle of leaning in, kissing, and then leaning back out before clicking and treating. It took a week or two of one session a day to solve the problem.
Like me, you may need to break your final goal into smaller ones. If you have a dog who’s already learned to bark at the mail, beg at dinner, or pull toward other dogs on walks, you’ll need to prevent him from continuing to practice those behaviors (by restricting access to or increasing distance from whatever triggers the behavior), and then start capturing whatever small step toward the desired behavior the dog can offer. If you need help, a good trainer should be able to help you make a management plan and plot out training steps that are achievable for both you and your dog.