Patricia McConnell, PhD, is an animal behaviorist and ethologist and an adjunct associate professor in zoology at the University of Wisconsin, Madison, as well as the author of numerous books on behavior and training.
Good Dog: Behavior & Training
Positive reinforcement is in the eye of the beholder
Visualize this scenario: Comet, a sixmonth- old Labrador in a beginner-level dog training class, ignores all distractions and charges back to her guardian, Barbara, after hearing “Comet, come!” As soon as she arrives, Comet receives effusive praise and petting from Barbara. A great example of positive reinforcement, right? Except that, in this very common situation, Comet responds to the actions of her guardian by turning her head away as though in disgust.
In this case, Barbara’s actions were a perfect example of positive punishment, in which something is added (pats and praise in this case) that decreases the frequency of a behavior (coming when called). Barbara couldn’t have done a better job of teaching Comet to ignore her if she had tried.
But wait — don’t dogs love to be petted? Don’t they love to be praised? Well, yes, they do. But then, no, they don’t. Lest you think I’m messing with your mind, put yourself in your dog’s position for a moment. Ah, yes, I know — being anthropomorphic can cause a lot of problems, but that doesn’t mean it isn’t useful sometimes. Bear with me here: Imagine that you’ve just done something wonderful, and your boss wants to thank you. She gives you a box of expensive chocolates, the kind infused with exotic flavors, like lavender and chili. You know she spent a fortune on them. Except, you don’t like lavender-chili chocolate; you like good old-fashioned milk chocolate from Hershey, Pa. Or, maybe you love flavored chocolate, but you’re on a diet, and being presented with a pound of chocolates is akin to torture.
The dog in the training class and you as an employee are in the same position: Someone is trying to give you “positive reinforcement” for a behavior they’d like you to repeat, but what they’re giving you is something they define as good and you define as, well, not very. Dogs may love petting, but not all versions of it, and not in every context. Praise may be music to their ears, but not all the time, or maybe they like funk and you’re singing jazz.
Many people know that using positive reinforcement (I’ll use PR from now on) is the most effective way to influence behavior, but knowing how to use it effectively takes more than remembering to carry cookies in your pocket. Without a doubt, one of the most common problems people have when using PR is not being thoughtful enough about what their dog defines as reinforcing. Of course, small bits of tasty food are fundamental for training a dog who truly loves to do what you ask, but they are not enough. You’re not always going to have Snacky Wackies in your pocket, and your dog isn’t always going to be hungry (well, that’s true for some of our dogs, anyway). When you do use food, keep in mind that the most frequent mistake beginners make is to use bland, boring stuff akin to a solitary oyster cracker. Good enough if you’re really hungry, but not worth leaving a plate of linguini with braised spinach and caramelized tomatoes for — or, if you’re a dog, a greasy food wrapper in the gutter.
But food, even great food, is never enough. The last thing you want is a dog who listens to you if you have edibles in your pocket and ignores you the rest of the time. And, although it’s hard to imagine if your own pup is a furry garbage disposal, there really are dogs who just don’t care all that much about food.
That brings us to petting and praise, which can also be great reinforcements if the dog truly likes to be petted (pats on top of the head appear to be as aversive to dogs as they are to people) and if it’s given at a time when a dog really wants it. An eight-month-old Lab/Husky cross is probably not going to be wild about pats on the head and hearing baby talk after he left his buddies in the middle of a wild play session. He doesn’t want to be petted when he’s in the middle of play any more than you want to have your back rubbed during an important business meeting.
Thus, food, petting and praise are great reinforcers if and only if they’re something your dog wants at the time they’re given, as opposed to something that makes you feel good for giving them. But there’s much more to PR than dog treats, pets and praise. This is where great trainers become truly creative, and is often the reason some dogs are phenomenally responsive. My favorite story of creative reinforcement comes from Bark blogger and CAAB Dr. Karen B. London. She was out with her newly acquired Lab mix, Bugsy, when they flushed a deer. She called him to come just before he sprang off in pursuit, and to her joy and amazement, he turned and ran back to her. Alas, before their outing, she had switched jackets on the fly and inadvertently left her training treats at home. She couldn’t even take off running to let Bugsy chase her because she was wearing cross-country skis and facing the wrong way in the ski track. Ever the quick thinker, Karen reached into her pocket, took out a used tissue and gave it to Bugsy when he arrived. Voila! Her dog was thrilled at this acquisition, and Karen’s creative thinking laid a foundation for a solid recall for years to come.
I turned to my own blog for examples of creative reinforcements, and my readers responded with enough ideas to inspire all of us. There were lots of examples of letting the dog have limited access to the very distraction she had been called away from (compost heap, dead bird, garbage bag). Many comments revolved around what trainers call “life rewards,” meaning the dog gets what she wants as a matter of course during the day (say, going outside after waiting at the door). Another easy way to reinforce your dog is to use a well-known cue as a reinforcer itself. For example, most dogs know how to sit on cue, and if you’ve used PR to train this behavior, the cue itself becomes a secondary reinforcer: “If I sit as asked I’ll get a treat! I love to sit!”
However, there are many more ways to expand your repertoire — you just need to know what your dog loves. Effective reinforcements include ice cubes, snowballs, bunny poop, cheering and clapping, pine cones, feathers, paper-towel tubes, digging, being allowed to sniff interesting things, running around the yard in crazy circles … and on and on. You can see that the list is almost endless — you’re constrained only by the number of things that make your dog happy and being mindful of your dog’s safety (e.g., you can sniff some things but not ingest them). One smart agility trainer had a dog who, no matter what she tried, would not go around the last weave pole. Then she remembered that her dog loved to jump into the car, so she set up the poles leading toward the car and used jumping in as a reward for weaving around the last pole.
Another agility trainer taught a fearful dog to enjoy the teeter board by using scent marking as a reward. Turns out the dog loved to lift his leg and mark objects, so every time he took one step closer to walking across the teeter, he got to lift his leg on the fence! This is a perfect example of what psychologists call the “Premack principle,” which says that a more-probable behavior can be used to reinforce a less-probable behavior. In other words, if your dog loves to dig, there’s a high probability he’ll do it if he has the chance. If that’s true, you can use digging as a reinforcer, secure in the knowledge that he must love it if he does it so often. If there is a low probability he’ll come when called after he’s seen a squirrel, you can use digging to reinforce a recall in that context.
Another PR tool that’s always handy to have is putting a problem behavior on cue and then using it as a reinforcer. My absolute favorite example of this came from someone whose dog loved mouthy play. The guardian put it on cue by saying “Rabies!” Now, she not only has control over the behavior, she can use it as a reinforcement. Brilliant! I did something similar (although nowhere near as funny) with Lassie by teaching her to jump up on me when I said “Be bad!” Lassie had come to me at 11 months as one of the worst leapers I’d ever seen. She’d launch herself up to a visitor’s face and, suspended midair, put her tongue halfway down his throat (we called her the French-kissing queen). It took many months, but I eventually taught her to sit while greeting company by using food treats and putting jumping-up on cue. She still loved to leap toward faces, so it turned out to be a great way to thank her for doing what I had asked, including sitting politely for company.
I’ll end on my current favorite example of creative thinking. One guardian had a scent-hound mix named Maggie who couldn’t be trusted off-leash. As happens to the best of us, the dog got out of the house one day and was just about to take off when the guardian spotted one of Maggie’s best friends close by. Nope, not another dog, but rather, a cat she loved to play with (a mutual pleasure, I’m happy to report). Our quick-thinking guardian grabbed the cat, held him up in the air and said “Maggie! Kitty!” and her dog came running. The three of them dashed into the house together, the dog and cat played joyfully, and the guardian breathed a sigh of relief. Now, that is quick thinking!
Granted, we can’t all grab a nearby cat, but I’ll bet there are at least five things your dog loves that you haven’t used yet. If you have an example of a creative reinforcement, share it with other Bark readers. You can send your ideas to email@example.com, or look for our Facebook page for discussion on this topic.
Culture: Readers Write
The science and art of naming your dog or Sometimes, what you say is what you get.
“His name is ‘Baby,’” Helen told me as she stroked her dog’s massive black head. “Baby” weighed in at about 65 pounds, was seven years old and had bitten 13 times. The last bite had been to Helen, when she tried to stop an attack on her disabled son. Needless to say, we had a lot to talk about, and one of the topics was her dog’s name. Helen explained to me that Baby had always been “her baby,” and that she did everything she could to make him happy. I countered, as graciously as possible, that Baby wasn’t actually much of a baby anymore. Rather, he was the equivalent of a 50-year-old man living in her house rent-free, not helping with the housework, and getting full-body massages on demand. Half jokingly, half not, I suggested that Helen change the dog’s name to something more fitting of his age and appropriate role within the family.
And that’s when I lost her. As soon as I mentioned changing Baby’s name, Helen’s face snapped shut like a book. Of course, we continued to talk through the appointment, but as I drove away, I guessed I’d never hear from her again, and I didn’t, even after calling her twice and leaving messages. In hindsight, I should’ve waited to talk about her dog’s name. Names are important, so important that Vicki Hearne wrote an entire book —Adam’s Task—about the weight of words in our relationship with dogs. What we call our dogs has meaning, and can have important consequences, both for ourselves and for our dogs.
One of the reasons that names are so important is the effect they have on us when we say them. Calling a male dog “Baby” makes it difficult to think of him as an adult dog, and makes it easy to excuse his behavior—it gives him “puppy privileges” that should’ve expired long ago. Labeling a Rottweiler “Brute” (as did one of my clients) does little to convince the neighborhood that your 85-pound Rottie plays well with Yorkies. Names evoke emotions in us, and those emotions influence our behavior. Since our behavior influences the behavior of our dogs and others around us, a name—all by itself—can have a surprising amount of power.
Emotions evoked by a name can have a profound effect even if you’re not conscious of it. Much of our behavior is driven by the unconscious—just look at the research of psychologist John Bargh, who found that people walk more slowly if you ask them to play word games with phrases that include indicators of age (like the words “wrinkled” and “bingo”). Believe it or not, if you’re named Georgia, you are more likely to move to the state of Georgia than you are to the state of Virginia, and vice versa. (To quote columnist Dave Barry, I am not making this up.) According to David Myers in the book Social Psychology, people’s careers are also affected by their names. Geologists and geophysicists are named George more often than is statistically predictable, and if you’re named Dennis or Denise, you are more likely to go into dentistry than if you’re named Tom or Beverly. Amazing stuff, yes?
Reflect, then, on the impact of naming your dog “Baby” or “Brute.” You say your dog’s name often, and the above-quoted research suggests that the repetition will have an effect. The good news is that the effect can be good as easily as it can be bad. I love spring tulips almost as much as chocolate (okay, not quite), and naming my huge white fluff-ball of a Great Pyrenees “Tulip” was one of the best ideas I’ve ever had. Just saying her name, “Twooooo-lip,” makes me smile. In a joyful swirl of classical conditioning, my love for her and for tulips have become intertwined in the best of ways. Surely Tulip is aware, either consciously or unconsciously, that her name, and thus she herself, make me happy—what a wonderful start to a relationship. Along those same lines, a friend of mine is considering naming her new dog “Sparkle.” After losing four beloved pets in the last year, she is more than ready to add a daily dose of light and joy into her life.
Thus, it’s useful to know that different types of sounds vary in their ability to get your dog’s attention. If you analyze the acoustics of spoken language, you’ll find that saying hard consonants, such as “k,” “p” and “d,” create what are called “broad-band” sounds, with lots of energy across a range of frequencies. If you were looking at a picture of the word “Kip,” you’d see a vertical spike (the broad band) for the “k” and another for the “p.” Those types of sounds are good at capturing your dog’s attention because they stimulate more acoustic receptor neurons in the brain than do the flatter sounds made by vowels and soft consonants. (That’s one of the reasons that clickers work so well—lots of broad-band sound.)
Thus, if you want your dog’s attention, you’re more likely to get it if she’s named Kip rather than Gwen. Of course, you can train a dog to pay attention to any sound at all if you condition her well enough, so if you want to name your dog Gwen, go right ahead. However, it’s useful, especially in performance events, to be aware of the effect of sound on your dog’s behavior. For example, short names with lots of hard consonants are great for people working dogs in fast-action events, such as agility and herding. The value of a short name is obvious: speed (you don’t want to be singing “Gwennnn-de-lynnnnn” when you’ve got a tenth of a second to get a response out of your dog) and focus (the consonants at either end of a name like Kip help you keep your dog’s attention). Indeed, so many working Border Collies are named “Hope” and “Jed” and “Drift” that conversations about the lineage of some dogs sound like “Who’s on first?” jokes. “Is your new little bitch related to Knox’s Hope?” “No, she’s out of McGregor’s Hope, sired by Jed.” “Is that Glynn-Jones’s Jed?” “No, I mean the Jed owned by….” And on and on. I’ve joked that for every 100 handlers in the sport, there are only 20 names for dogs.
That said, I must add that there’s something satisfying about a two-syllable name; “Pixie,” “Tulip” and “Sparkle” all flow off the tongue in a way that just feels good. I’ve also wondered if, in some cases, two-syllable names can actually help get a dog’s attention, in that the first syllable acts almost as a primer for the second. Perhaps the handiest names are the ones with a lot of flexibility. My forever dog’s name was Luke, but his recall signal was his name said twice: “Luke Luke!” When we were working sheep and the pressure was on, I’d belt out “LUKE!” to bring his attention back to me. In quieter times, if he did something silly, I’d say, in a rising, drawn-out drawl, “Luuuuuu-cas, what are you doing?”
Luke’s name brings up one more thing to think about when you’re naming your dog (and yes, of course you can change a dog’s name if you don’t like the one she came with!). I named Luke’s daughter “Lassie,” not because of the acoustics of the name, but because she came to me as a dog rejected by two people who had missed her potential to be as devoted and responsive as the television star of the same name. But listen to the consequence of that choice—say the names out loud: Luke. Lassie. I gave two dogs in the same household names that start with the same sound, and as hard I tried to keep things clear, it made life a little bit more confusing for the two of them. You can see it for yourself in a video I made, Feeling Outnumbered, in which I tell all my dogs to “Wait” at the door of the car, and then release Lassie by calling her name. If you watch carefully, you can see Luke start to move forward when he hears the “L—,” and then self-correct when the rest of his daughter’s name comes out. Luke and Lassie were so amenable and responsive that my mistake barely mattered, but keep this in mind when you’re naming a new dog. I sure will. Living with humans is confusing enough for dogs, why make it any harder?
In summary, there are several facets of a dog’s name that bear consideration. It’s good to be informed about all of them, but I have to admit: When push comes to shove, I’d vote every time for the name that fit my dog’s personality and that made me happy to say over a name with the “proper” acoustics. It’s good to be aware of all the ways a name can affect your dog’s behavior, but nonetheless, a dog by any other name … will still roll in cow pies.
Copyright © 1997-2016 The Bark, Inc. Dog Is My Co-Pilot® is a registered trademark of The Bark, Inc