Be Careful What You Teach Them …

When I was first learning to train service dogs, my instructor taught me that, once you put something on “cue control,” that is, teach the dog to do it when asked, the dog will no longer do it unasked.

Those of you with dogs and a little training experience, stop laughing.

This is one rationale behind teaching dogs to “speak” — bark on command.

Problem is, no one told the dogs about this bit of learning theory.

There is a key caveat to this piece of wisdom: The dog will no longer do it unasked unless he really, really wants to.

Bonnie Bergin (my teacher) discovered this one morning, when a particularly rambunctious group of adolescent Labrador service-dog trainees learned about tugging open a refrigerator door. One day, the ringleader of the litter, Xavier, let himself and some buddies into the training room (opening the door from the outside yard), tugged open the fridge, and helped himself to a large hambone.

The ensuing melee was quite dramatic. None of those Labs became service dogs.

I learned about the “unless they really want to” caveat to the lesson about dogs not volunteering named behaviors too late: I had already taught Jana to speak on cue. She uses her words. A lot. She’s quite opinionated, in fact.

Cookie, please

But I digress. The latest example of dogs doing unbidden that for which they have been amply rewarded in the past involves my shoes. I recently wrote about how eagerly Cali and Jana bring my shoes or sandals when it’s time for our morning walk.

Well, now they bring my shoes and my sandals. And a pair of slippers or flip-flops too. I reward the correct two shoes and wordlessly return the others to their proper place, hoping to extinguish this behavior. Not only is the flood of shoes not diminishing, it’s extending beyond walk time.

I occasionally turn around to find Jana, hopeful look on her face, shoe in mouth, standing behind me … at any hour of the day or evening. When I ask for anything — bowl, leash, collar, toy — Cali will often dash off enthusiastically … and return with a shoe.

On the bright side, they are both really good at “getting the other one,” so I never need to worry about being barefoot or mismatched.

So, be careful what you teach your dog; she might be smart enough to turn it into a game you never anticipated.

What Is Dognition?

Jana dognition
In last week’s post, If Cali Were Brian Hare’s Dog…, I mentioned Dognition. What is it?
First: full disclosure. Jana and I were among the beta testers of the site, for which we received no pay, but Jana did receive the bandanna she’s shown wearing above. This is not a paid post or advertisement. It is my opinion, based on my experience with the site.
Brian Hare, whom I wrote about last week, is the co-founder and chief science officer. Dognition is a site that studies canine cognition through a series of games and experiments that site members do with their dogs, at home. The Dognition team includes some innovative canine cognition researchers. It is a clever way to gather a huge amount of information about a large variety of dogs. It’s also (usually) fun.
The initial set of games and exercises creates a “profile” of your dog. There are nine profile types, ranging from the “Ace,” a problem-solver with strong communication skills, to the “Protodog,” who, according to the brief description on the site, is “reminiscent of the first dogs that began their relationship with early mankind.” I am guessing that that’s a nice way of saying “can’t follow human gestures.” The writers on the site excel at couching bad news (“your dog flunked this test”) in very positive terms.
Renaissance DogJana is a Renaissance dog, good at a little bit of everything. I never finished the profile exercises for Cali, though.
The reasons are, I think, the site’s biggest drawback: Nearly all of the games require two people, and they are very hard to do alone. The site steps you through each set of exercises, and you have to enter information about your dog’s response before getting the next prompt. Another drawback (especially for multiple dog households) is that the exercises are repetitive.
I get it. They want data that they can use. They usually have a baseline exercise, which you do a few times, then the real one, which might have a few variations. For example, you might put a treat under a cup while showing the dog where it is. Then you might introduce a second cup, but still show the dog where the treat is. Then you might pretend to put treats under both cups but point to the right cup. Then you have someone hold the dog where she can’t see you and you hide the treat and point to the correct cup … you get the picture. You do each step three to five times. Jana can lose patience. When I am testing one dog, the other rapidly loses patience as she waits (and waits, and waits) for her turn. And I don’t have a second person here to cover Jana’s eyes or whatever else is needed.
So, those problems aside, I have learned a lot from Dognition. The site includes lots of articles and information on canine cognition. They do a good job of explaining the science behind the games and the profiles. And they send occasional tips, games, and suggestions that can be done by one person. I’ve gotten some great ideas for games to play with Jana and Cali and experiments to do with my students and their dogs. And I enjoyed finding watching Jana’s progress through the exercises. I’m curious about how much usable data the Dognition team has gathered and how they are using it.

If Cali Were Brian Hare’s Dog …

photo 4If Cali were Brian Hare’s dog, or rather, if she had been his dog years ago when he first started studying human-animal communication and animal cognition, there would likely be no Duke Canine Cognition Center, no Dognition …
I am not saying that Cali is not intelligent (far from it!). Before I get into that, though, some background:
In Brian Hare’s (wonderful) book The Genius of Dogs, he tells the story of how his interest in canine cognition came to be. Hare, now a leading researcher and innovator in canine cognition, was studying primates. As a college student, Hare participated in a study that investigated whether bonobos and chimpanzees understand the intention behind gestural communication. The chimps flunked the test. But, while conducting this research, Hare commented to his professor that his dog could do that. And history was made.
Hare and his professor found that dogs, even very young puppies, understand the communication intended by gestures such as pointing. Chimps must be taught, and, even then, rarely generalize. Wolves don’t do as well as dogs, even if they’ve been raised by humans.
In the first test that Hare did with his dog, Oreo, Hare threw three balls into a pond. After Oreo had found the first one, Hare pointed to the second, then the third. Oreo effortlessly located the balls by following the pointing.
Simple, right? Your dog can do that, I bet. Most dogs can. Jana can. Alberta can. Cali? Not so much.
Cali loves to play ball. She gets very excited when we’re playing, and often starts running out in anticipation of my next throw. She gets so excited that she neglects to keep her eye on the ball … and often ends up with no idea where the ball landed.
So, comfortable in the knowledge that dogs can do this, I point. And she invariably runs off … in a completely different direction. We play the “hot, cold” game. I pretend to throw the ball again, waving the Chuckit in the correct direction. To her credit — and showing considerable intelligence — Cali is never fooled by this gesture. I point some more. I walk in the right direction. I do everything except point a neon arrow at the ball.
Meanwhile, Cali continues running huge loops in completely wrong directions. Eventually, she ends up in the right section of the field and, using her excellent nose, locates the ball. She hardly ever loses her ball. But she just doesn’t get the pointing thing.
According to the studies I’ve read, Cali performs about as well as an unsocialized wolf. So what does this mean?
Aside from the obvious — that if Hare’s dog had performed as dismally as Cali, the science of canine cognition would have never been born — it means that Cali lacks this particular type of social intelligence.
She brims over with other types of social intelligence, though: She is extremely empathetic and affectionate; she is overwhelmingly friendly and nonjudgmental; and she is playful and happy. She also has a great memory and can use her nose to find hidden items in seconds flat (unless the item is a tennis ball nestled in the grass, of course).
The point is that dogs, like people, have different types of intelligence. Each individual excels at some things while faring more poorly at other skills. And that’s just fine.

Is Your Dog Smarter Than A …

Is your dog as smart as a human 2-year-old? A 5-year-old? A (gasp) teenager? Does it depend on what breed your dog is?
We can’t help it, we humans. We want to put everything into neat little human-constructed boxes. That is, I think, what is going on when people try to define dogs’ (or other non-humans’) intelligence in human terms. That and the common, if arrogant, human assumption that we are the smartest creatures, so everyone else — dolphins, dogs, starfish — can and should be evaluated, based in how they compare with us in human-like ways.
But really, how many human 2-year-olds would you trust to guide you across a busy street? Or turn loose in the wreckage of a natural disaster or terror site, with the expectation that the little tyke would let you know where the survivors are trapped? We use dogs to find lost 2-year-olds, don’t we? And protect them (and other humans) from diabetic coma or severe peanut allergies, warn of their impending seizures, coax those who have autism or have suffered trauma to connect — and so, so much more.
The basis for comparison is obviously flawed. Dogs are much like human toddlers in many ways, it’s true — their unbounded love of play; their sweet willingness to befriend just about anyone. Yet they are so much better at some things than any child could ever be — better at some things, such as anything based on scent, than any human of any age could ever be.
So, how should we measure, evaluate, understand canine intelligence?
We can start by acknowledging that intelligence is a complicated concept — there are many types of intelligence. Among people for example, there is social intelligence or emotional intelligence, there is numerical or problem-solving or analytical intelligence. Business acumen, logic, performing well under extreme stress — all of these might be considered different skill areas or types of intelligence. Intelligence is what helps you (or your dog) navigate life, with all the challenges and detours it throws in your path. We are all stronger in some areas, weak or ridiculously incompetent in others. The same is true of dogs.
We can also think about the skills that dogs have that have no parallel in human ability or intelligence — and the myriad ways we can help dogs develop and use those skills in partnerships that make life better for humans and dogs.
Some dogs excel at reading people’s body language. According to several prominent dog cognition researchers, among them superstars Brian Hare and Adam Miklosi, dogs — even very young puppies — excel at reading humans’ pointing gestures and where their humans gaze. This is a type of social intelligence. I am sure that many, many dogs excel at this. However, not all dogs do. I know. I live with one who fails miserably at reading gestures.
Other dogs (including the one who cannot follow a pointing finger to save her life), can intuit a person’s mood and provide exactly what is needed: comfort, humor, affection, appeasement, a favorite toy.
Still other dogs are great problem-solvers. They analyze each new situation and map out a solution.
Some dogs are born to … fill in the blank: Provide mobility assistance, search out bombs or drugs, find lost or hurt people, detect tumors, comfort lonely elderly people, make children laugh.
I don’t think it matters whether your dog is smarter than a toddler. I don’t think it is a fair or relevant comparison. What does matter is assessing each dog’s strengths and weaknesses, his or her specific areas of intelligence. Then, we can figure out how to stimulate and challenge each dog in the ways that will allow him or her to succeed, thrive, and enjoy life to the fullest.