Guest Posts
Print|Email|Text Size: ||
Home-Schooling for Dogs Could Be Catching On
“Do As I Do” scores high
Claudia Fugazza

A rambunctious five-year-old Labrador Retriever who until a few months ago knew not a word of any language, obeyed no command, charged around the house or zipped through any hole in the fence before one could utter the name he didn’t seem to recognize has become my 91-year-old mother’s great and constant companion. He sits or lies by her when she is sitting or lying down. He moves with her when she goes somewhere with her walker and when she tells him to give her clear passage. He accompanies her when she walks around the pool for exercise. She says, “He is a good boy.” My mother has never trained a dog. She had a nice trained dog once, but she had been trained by someone else and given to her.

But Rocky, as he was named by my mother’s granddaughter, received no formal instruction from any source. He was neutered, which helped slow him down, but more profoundly, he and she opted for companionship and accommodation over ignoring each other. She talks to him constantly, telling him what she wants him to do. If she praises him, she is not effusive. She may occasionally slip him some food when she is cooking, and he will if given a chance steal her breakfast bagel. There is no system to it, but there is consistency.Top of Form

More than a few dog trainers who follow behaviorist principles that require a stimulus, a reward or punishment, for learning to occur would argue that Rocky is untrained—that is that he still will not perform on command the actions demanded of him—except he comes when called. He moves when told. He tells my mother when someone is at the door and stands by her when she opens it, thereby providing at least the illusion of protection. If that is not training, what is it?

My friend and fellow Psychology Today blogger, Marc Bekoff (“Animal Emotions”), might call the process dog teaching or dog learning.

It might not be as quick or as systematic as one of the common schools of training, including those that use electric collars and choke chains and those that rely on clickers and food rewards or other positive re-enforcers. But then again the results might be quicker, deeper, and longer lasting.

I have seen no statistics on the numbers of dogs educated in this fashion, but I imagine it is substantial. Essentially it relies on the dog’s innate curiosity, desire to please, and recognized ability to imitate behavior and recognize words and emotions, traits which arguably thousands of years of living with humans have served to enhance. It also requires the human have an interest in being with the dog and interacting with him or her in a meaningful way—what used to be referred to as “quality time” with the hound. Praise and rewards are meted out more according to the person’s nature than any program or schedule. They do not have to involve food. Our Kelpie Katie was unmotivated by food—she would ignore food rewards—but when a tennis ball appeared she went on high alert. Even then the ball was not essential to her learning something.

This intuitive style of dog teaching is not without its intellectual underpinnings thanks initially to Edward Tolman in the first half of the last century. He proposed that learning had intrinsic value and that people and animals could learn in the absence of immediate rewards—latent learning it is called. That idea underpins what is called the social theory of learning, which also views learning as a social endeavor that can involve imitation of behavior that is demonstrated or verbally described.

In an article in the January 28, issue of Applied Animal (Behaviour Science, entitled “Should old dog trainers learn new tricks? The efficiency of the “Do as I do” method and the shaping/clicker training method to train dogs,” Claudia Fugazza and Ádám Miklósi of the Eötvös Loránd University in Budapest, Hungary, look at a canine system of social learning that relies on the dog’s great capacity for imitation called Do As I Do (DAID) compared with clicker training, which relies on the timely delivery of rewards to employ the dog’s associative abilities in shaping its behavior. (The article is only available by subscription, but here is the Abstract.) The clicker becomes a stand-in (secondary re-enforcer) for the actual re-enforcer, usually food. Clicker training is individualized instruction that requires the dog to figure out what earns rewards.

Fugazza, a graduate student in ethology developed Do As I Do in order to study social learning in dogs. To do that she had to develop protocols for teaching them. Judging from its success, it should gain a wide following. In this method, trainers, usually the dog’s primary human companion, use standard reward-based techniques to teach the dog to associate a small number of gestures with the command, “Do It!” The dog is then shown a new task and taught to perform it upon being given that command.


For this study, Fugazza and Miklósi compared the speed in learning three sets of tasks of increasing complexity, from knocking over a glass (simple) to opening or closing a locker or drawer (complex task) to a sequence of actions, like hopping on a chair and ringing a bell or opening a locker and removing a purse (compound). Objects were involved in each task that were not considered part of the family dog’s normal repertoire so that mastery of the task could be construed as learning. In the simple task there was no difference in performance between clicker-trained dogs and Do As I Do dogs, but that changed as the tasks became more difficult. Do As I Do dogs performed noticeably better, with more of them learning the task in the allotted fifteen minutes than clicker-trained dogs. 

No one knows how the dogs are making the connections, and in their conclusion Fugazza and Miklósi thought it more important to downplay that result in favor, Miklósi said in an email, of providing trainers with as many methods as possible so they can choose the one best suited to their needs.

That is a tactical decision rather than a scientific one. It is grounded in the recognition that, especially commercial dog trainers and trainers of working and service dogs, like to use what has worked for them in the past with the kind of dog on which it has worked. That is one reason punishment-based forms of dog training persist. 

For home schooling, time, patience, devotion—and a daily reminder of who has the big brain—are the keys to success and those come from discipline we often need more than the dog.

Used with permission of Mark Derr and Psychology Today, see more from Mark Derr’s blog “Dog’s Best Friend.”

Also see http://thebark.com/content/dogs-are-asked-just-do-it




Mark Derr is the author of A Dog's History of America, Dog's Best Friend, The Frontiersman: The Real Life and Many Legends of Davy Crockett, Some Kind of Paradise, and numerous articles on science, environment and transportation
CommentsPost a Comment
Please note comments are moderated. After being approved your comment will appear below.
Submitted by Lisa Brown | March 8 2014 |

Thank you for this wonderful article. For years friends had blank stars when I've told them that we and our dogs have an understanding. They stop when we say "whoa", we use a clicking of the tongue and a pull on the leash to go right or left and so on. We've been kindly told that we are not "doing it right." Thank you for showing that like people, dogs learn in different ways.

Submitted by Carol | March 10 2014 |

This describes perfectly what happens in our multi dog household (we have always had a "pack" over the last 40 years). I would have to say the older dogs have always provided a very strong "do what I do" lesson for younger pack members, probably even more influential than anything we do. I think this is the first time I've heard described "family expectations" for dogs. It works!

Submitted by Brenda | March 12 2014 |

Finally, an article that describes exactly how I have "trained" my dogs for the past 19 years. I have never taken a dog to an obedience class, and I never did take to the clicker training method - I never understood why you would do both-click AND reward, why not just reward? All of my dogs (all adopted & all but one were at least a year old) have simply learned to understand what it is I want from them. They are treated as "equals" and the bond we form is incredibly deep.they want to please me & I, in turn, want to please them. My husband has always laughed when people ask how we train them, he says it's just my innate ability to bond with our dogs, to understand what they need. They have a tremendous amount of freedom to do what they want, but they know when I need them to do something. I truly believe that before training a dog (through any method) a person must really bond with that dog.

Submitted by Selena Dwyer | March 20 2014 |

The reason you click and reward is that the click is a marker. For example: If you were teaching a dog to sit and by the time you rewarded him his butt had left the floor, he might think he was being rewarded for standing. By using the clicker you can capture the instant his butt touches the floor, like a snapshot, telling him 'that right there is the behaviour you are being rewarded for'.
An example from training with my own dog: I was asking him to walk close to me as we walked lengths of the room. At the edges I would ask for a sit. Initially (I believe we had been doing some 'down' training prior) he kept going into a down from the sit. The first time I clicked as his butt hit the floor and rewarded even though he was lying down by that point. The second time I did the same and he stopped midway to lying down. The third time he did it perfectly:)

Submitted by Josh W | March 13 2014 |

Not sure I understand the comparison between "clicker" training and DAID. Clicker training isn't just shaping as it looks like what the study compared. The videos appear to show the trainers eliciting (cueing?) behavior by doing it first, the dog performs the behavior, click, reward. Seems like pretty straight forward positive reinforcement to me. Now it is a novel way of getting the behavior to begin with but not a great comparison to "clicker" training.

Submitted by Lori Kline | June 4 2014 |

I agree. I think once the dog understand that "Do it" means to mimic the human, I can see how it could be faster than clicker training but it seems to me it falls under the same rules. Still fascinating. And yes, dogs do learn by observing others (I think there were some American studies on this using dogs observing other dogs. Can't remember what the results were but I believe they did find dogs could mimic but a primate couldn't or something like that.)

More From The Bark

More in Guest Posts:
Time Magazine and Designer Dogs
The Difference Between Guide Dog Breeds
Spice's Amazing Transformation
Career Moves
Timmy's Amazing Transformation
Learn How To Train Dogs at ClickerExpo 2015
Defusing Awkward Situations
From the Streets to the Gallery, All Thanks to the Dog
Jedi Surfs
This Dog Loves Guitar!