Doggy generalists and doggy specialists (part one)

Cali, a golden retriever, sits in the kitchen doorway

I’ve recently heard not one but two experienced trainers, trainers I admire greatly, say that dogs don’t generalize.

What?!

The most recent incident was in a training class. The trainer was talking about practicing at home and said that dogs don’t generalize as a way to explain why the dog’s reaction to a cue or a piece of training equipment might be different at home than at the training center.

Generalization, in a dog-training context, means that the dog can recognize a cue in different environments. Some examples:

  • Your dog has been to your  neighborhood dog park a bazillion times. She’s been to the dog park near your parents’ house a half-bazillion times. When you take a road trip and stop at a dog park you find near your hotel, she immediately recognizes it as a dog park and runs off happily to play.
  • Your dog has a toy box. Your sister’s dog has a toy box. Your parents’ dog has a toy box. Your dog happily helps herself to toys from the toy box at each home. You visit a dog-owning friend for the first time. Your dog homes in on the toy box, grabs a toy, and the dogs start playing.
  • Your dog knows to sit quietly and wait while you prepare her dinner, waiting until you say, “OK” before gobbling her food. She does this whether you’re at home, in a hotel, at your mom’s, or anywhere else.
  • You’ve taught your dog not to jump on visitors. A new visitor comes into the house. Your dog has never met this person before, yet she knows not to jump on him.

I could go on, but I am guessing that you get the picture.

Of course dogs generalize. They could not possibly learn otherwise. No one could. There is simply no way to create and practice every single scenario where the dog will be asked to sit or lie down or wait for food or not jump —  or do / not do anything you might imagine.

Dogs learn to generalize as they learn to learn — with repeated practice. With a new puppy (or newly adopted dog of any age), you start teaching the dog house rules. You also (one hopes) teach basic manners, things like not jumping on people, not mouthing people, not barking every time the next-door neighbor walks past the window. Then, other members of your family ask the dog for the same things. And visitors do. You ask for them when you are in the kitchen, the living room, the basement, and the back yard. And on a walk or a road trip.

Your dog learns, after experiencing this phenomenon a few times, that the cue applies everywhere. The dog then generalizes that knowledge to other cues. Verbal cues: “Sit,” or “Quiet.” Visual cues: the toy box or dog park. Routine cues: you putting on the shoes you always wear for walks or closing down your laptop at 5 pm. These cues, and similar cues, take on meaning that applies even if you are not in the same room or you’re visiting someone else’s house or the toy box looks and smells different.

By the time you’ve taught your dog a few cues, she’s figured out generalization.

You can also teach your dog to specialize, which I will talk about in next week’s post.

 

The saddest sounds

10-week old Cali, a golden retriever, lies on a brown dog bed
Don’t leave me …

A recent Bark column muses on humans’ susceptibility to manipulation by dogs. Specifically, by the sounds they make in sadness. Sadness that occurs only because we humans are not meeting their expectations.

Boy do I know how that works.

When Cali was a tiny pup and Jana a beleaguered 8-year-old with a new baby sister, I made a point of taking Jana for a (very short) solo walk each day. This was partly to get Cali used to being alone briefly. The first time we did this, within seconds, the saddest, most mournful howl I have ever heard wafted out through an open window. I was probably a whole 10 feet from Cali but, you know, there was a wall in between.

Cali has deployed this mournful howl a few additional times over the years. (She’s 7 now.) She’s added to her repertoire, too. She has a range of sounds, including sighs, snorts, scowls (ok, those are silent), exasperated exhalations, grumbles and mutters under her breath, and more. And, yes, a whine. It’s a tiny whine, very soft and short. It’s also very, very sad. Heartbreakingly sad. This whine is used only when Cali is outside and wants to come inside, and no one is there to make the door magically open.

This, naturally, happens only when Cali has refused to come inside despite being offered several opportunities, and I have given up(!) and gone upstairs to work. Within oh, about 3 minutes, there’s that tiny whine. I could easily miss it but somehow it penetrates whatever fog of concentration I am in. When I go back downstairs to let her in, Cali is always happy, relieved, and reproachful, all at once.

I’m not the only one to be expertly and repeatedly manipulated by a sad dog.

My doggy cousin, Jaxson, has created a magical combo, a unique whining sound plus guilt-inducing look, that gets him the most coveted seat in the house: Literally in between his mom and dad. The one space on the sofa he’s theoretically (very, very theoretically) not allowed. There’s nothing unique about dog whines, of course. Whole orchestras could be woven out of different dog whine. Jaxson’s whine is unique in that this specific note is deployed only when he’s on the sofa but not between them. That is, only one pair of hands can reach him to pet him and only one person’s attention is focused on him. The unique sound effectively terminates this intolerable condition.

The Bark column mentions research that found that humans with pets are more susceptible to animal distress vocalizations than other people and that “dog whines sounded saddest of all, and sadder than cat meows.” Other research has found huge changes in canine vocalizations as a result of their domestication. Sure. They’ve got our number. They’re pulling out all the stops in their quest for the upper hand … er, paw … in the household.

 

Dogs and Deer

The house where we are dog-sitting is above the Bitterroot Valley

We’ve been hanging out in the Montana wilderness, dog sitting. Well, the edge of the wilderness, anyhow. And it’s fawn season.

We were out in the play yard with a motley collection of tennis balls, most of which were cleverly camouflaged in the grass. We do this three or four times a day — take three balls and three dogs. Play for a while. Return to the house with one or, if we’re lucky two balls …

I hadn’t seen many deer around the house at all, and none in this fenced play area, and I hadn’t seen a single fawn yet this year. So I wasn’t thinking about deer as I walked around looking for lost tennis balls. We were down to one, which Cali was carefully hoarding.

Suddenly, I saw a flash of brown and white. A tiny fawn nestled in the grass. I’d startled her (or him). The fawn ran. The dogs, being dogs, noticed and ran after. I yelled, dogs chased, things got scary and noisy. The fawn got to the fence … and tried to get out. Non-trigger warning: No one gets physically hurt.

This part, though terrifying, was also very interesting. Cali got to the fawn first. Tail at half mast, wagging, she sniffed. She did that “hold back and stretch forward at the same time” thing she does when she’s nervous but her curiosity mostly overcomes her apprehension. I yelled at her to get away. She did.

Then Mack got there.

She looked more serious, and I screamed at her to get away. She’s very obedient, so she did. But … the fawn was scared and couldn’t get through the fence and started bleating. With every bleat, Mack returned in a flash. I’d yell. She’d leave.The cycle would repeat.

By this time, Cali had gone back for another sniff. I also kept telling Cali, “NO!” and she’d look at me, then sniff and wag some more.

I finally got there (this all happened in about 15 or 20 seconds …) and grabbed both dogs’ collars. I dragged them away and … Alberta sauntered over to see what was happening. I called her, too, and she came right away. Good girl!

After dragging the dogs into the house, I went back to see if the fawn was stuck. She was gone. I really hope her mom came and got her, but I can’t really know whether she’s OK. It’s been about a half hour, and I am just now starting to breathe normally again. My heart is still pounding, though not quite as fast. The adrenaline is subsiding, I guess.

The dogs were doing what dogs do. Which is a problem where deer are also doing what deer do.

I was happy to see how gentle Cali was, but the fawn didn’t really see that and was justifiably terrified. I’m less thrilled that Cali did not come when I called her.

I’m not as confident about the beneficence of Mack’s motives, but I am grateful that she listened to me (multiple times) though less happy that she kept going back into the fray.

I’ll go out to the play yard and make sure it’s deerless before taking the girls there again but … I’ll also be happy to be back home, where the neighborhood deer respect our 6-foot fence and stay out of Cali’s yard.

Your Dog Is a Logician

Koala, a black Labrador, rests. She's wearing her guide harness.
Koala loves to cuddle

Guest Post by Deni Elliott

Guiding Eyes Koala and I have few conflicts, but those that we have get tiresome for the person-side of the partnership. Licking. Koala is a licky dog. I don’t like to be licked. I think I have found a solution to the licking conflict and realized something more about canine language comprehension.

I’m not unreasonable. A quick kiss now and again is appreciated and fair exchange for an ear scratch or a snuggle. But, spare me what became a nightly routine: The enthusiastic leap when I invite Koala to come up on the people bed for a cuddle, ending with her upper body on my chest and nose at my chin. “Oh, you let me up on the bed!” her body language says. LICK. LICK. LICK. “I’m so happy to be up here!” Rapid tail wag. LICK. LICK. LICK. “I’m such a lucky dog!” LICK LICK LICK. My plaintive, “Enough!” or commanding, “Off the bed!” ended the licking, but ignored Koala and her intent. This was not how I wanted to end the day with my canine partner.

Recently, I invited her up on the bed and, as she came close to my face, I calmly and nicely said, “If you lick, you’ll need to get off the bed.” Koala stopped. She lay down next to me with more restraint than usual, nose close to my chin. Her tongue slowly reached out to touch my face. Again, I said, “If you lick, you’ll need to get off the bed.”

She stopped. Sighed. Relaxed against my side so that I could stroke her head. Soon she shifted to her back so that I could rub her belly. No licking, just a peaceful, happy dog. A few minutes later, I said that it was time for her to go to her own bed, which she did without protest and without being rejected. I heard Koala settle down on her dog bed next to her Golden Retriever sister and thought about why my warning worked.

Conditional reasoning starts with compound sentences that use “If, then.” Dogs know the “if, then” construction. Sometimes the conditional is time, such as when Pam says, “We’ll go for a walk, and then I’ll give you dinner.” Koala and Cali know the concepts of “walk” and “dinner,” but on hearing this sentence, they head for the door, not the dinner bowls. More often the conditional is action: “If you get the paper, then I’ll give you a cookie,” “If you sit quietly for a few minutes, you’ll get your dinner,” “If you come here, then I’ll rub your ears the way that you like.” The “if, then” condition sets up a trust relationship between dog and human. Dogs that stop coming when called do so because they think that their humans have fallen down on their part of what should happen when they obey.

Koala’s understanding of the “if, then” connection when I said, “If you lick, you’ll need to get off the bed,” was even more complex. This sentence had a condition — If you lick — but a negative consequence. Koala needed to turn the sentence around to reason, “I don’t want to get off the bed, so I better not lick.”

And even more complicated was her realization that she’d have to get off the bed in a few minutes anyway to go to her own bed and that that part of the nighttime routine was not a punishment. The question was whether she could control her licking so that she got a cuddle on the people bed for a few minutes first.

Koala’s success was evidence for her ability to master human language and what is implied by what her people say. Its mastery that all of our dogs are capable of achieving. The limitation is with us humans, who often fail to see how asking our dogs to use their conceptual abilities can make life easier for both humans and canines.

What It’s Like to Be a Dog

Cover of Gregory Berns's book What It's Like to Be a DogI’ve had a serious crush on Dr. Gregory Berns ever since he published his first MRI studies. Those showed that dogs’ brains’ pleasure centers light up when they catch the whiff of a beloved human (or dog). There’s so much to love about his papers and his book How Dogs Love Us. So I was really excited about reading his newer book, What It’s Like to Be a Dog.

It’s well worth reading, and I enjoyed it. But … it wasn’t what I was expecting. There’s some really cool stuff, like the explanation of how dogs’ brains look when they’re doing the equivalent of the Marshmallow Test. I’ve played around with that a bit with Koala and Alberta, though I lack access to an MRI machine. So I was very interested in his findings. It turns out that some dogs do well with deferred gratification and others … not so much. You might notice that I haven’t talked about doing a marshmallow test with Cali. I don’t need a fancy machine to tell me that she lacks impulse control.

I was a little disappointed with some of the detours from living dogs’ brains into the long-ish discussions of the brains of deceased seals and Tasmanians. And I was distressed by the chapter on dogs and language.

I know that any sentence that pairs non-human animals with language raises the hackles of many people, scientists and non-scientists alike. I also think that there are many, many definitions of language and that dogs, particularly those with close human connections, understand a lot of what we say and do and they communicate with us in sophisticated ways. Lack of understanding of their “language” does not diminish its value. I get irritated when people choose a very narrow, very human-centered definition of language, such as one that is focused on semantics and grammar and written representation of a language, and then say, ‘see, only humans do this so only humans have language.’

Dogs communicate. They use their whole bodies — ears, tails, hackles, eyes, facial expressions, as well as scent and sound, to communicate. And dogs excel at reading the nonverbal communication of other dogs, humans, and often of other animals like cats. Other non-humans do this as well. Dogs are able to read humans far better than humans can read humans.

And dogs understand a lot of what we say to them. They might be assigning meaning to a combination of words and body language cues to understand our feelings, our desires, our mood rather than attaching the specific meanings that we do to individual objects or concepts. While I don’t expect Cali to speak to me in English or read the newspaper, much of the communication that I have with Cali — and especially what I had with Jana — is clear and meaningful.

Berns’s discussion of language, how he tested dogs’ understanding of words, and his interpretation of those results are very, very human-centric. He talks about the mirror test, which I believe is not a fair test for dogs. His comments on dogs’ lack of a sense of self or others: “My beloved Callie probably didn’t have abstract representations of me or my wife or my children. No, I was just that guy who feeds me hot dogs …” are off-base.

Dogs’ sense of self and others is primarily rooted in scent, not sight or sound. Berns himself showed that dogs recognize the scent of family members and respond differently than to the scent of unfamiliar humans or dogs. So I was mystified and saddened by what felt like a dismissal of the individuality of dogs’ selves and their relationships with key humans (or non-humans).

Despite a few disappointing chapters, I do recommend the book. I the insights into how dogs’ brains work are fascinating, and even where I disagree with Berns’s conclusions, I enjoy learning about his research and his understanding of dogs. Dr. Berns is still my favorite neuroscience researcher, and he’s a great writer. Check out both of his books if you haven’t already!

 

Guide Dog Haiku

Deni Elliott learns to work with Guiding Eyes Alberta, who is now retired.

Several Guiding Eyes dogs’ human partners recently posted haiku and other poetry to a graduates’ email list. The poems show their appreciation and love for their guides. A member of the list asked (and received) permission to share some of the poems, which appear below. Feel free to add, in the comments section, your own service- guide- or pet-dog haikus, odes, ballads … or tributes in any literary form.

Naughty puppy face
Harness on, working face on!
What to do without?

Night comes, harness off
Naughty puppy face once more
We dream together.

 

The trees and sky breathe
My golden girl goes forward
Our hearts together

 

My vision as wide
As a dog can see, hear, smell
Guiding Eyes radar

 

Walking by my side
You safely show me the way
Teamwork every day

 

Our talks as we walk
Open volumes clearly spoken
Unheard by strangers

 

They don’t know our language
We speak silently yet so loudly
RIGHT!  LEFT!  STOP!  I LOVE YOU!

A movement, a language, a laugh
in voices so clear to us
So invisible, so silent to strangers
Roxanne, I hear you

You speak more loudly
“You do, too, when you smile at me.”
I smile back
A wag of  tail
A snort and shake of collar
A lean against your leg
A huff, a snort.
I smile back

Strangers never know
We laugh out loud at them
Out loud but silently
Our talks when we travel
Volumes never heard so clearly spoken
So secret, so open

 

The partnership and communication between guides and their humans is unusual, but service-dog partners and working-dog partners often experience  a comparable connection. True communication develops best in relationships where both partners’ roles are recognized and each acknowledges the necessity and the significance of the other’s contribution. This idea goes to the heart of the Thinking Dog Blog and my reasons for writing it, which is why I wanted to share these heartfelt tributes to guide dogs, both working and retired.

Where’s Cali?

Cali hides in tall grass

Cali has started a new game. When she doesn’t want to leave someplace (Jacob’s Island Bark Park in Missoula, for example), she hides. She’s really good at hiding, too. See if you can spot her in the photos.

I could get angry with her, say she’s misbehaving, that she’s a willful adolescent. I could also admire her intelligence, strategizing, and ability to read my intentions.

It takes a lot of thought and awareness to hide. Her ability to do so shows that she can think about what I can see and figure out where she can go that I won’t be able to see her. She can anticipate going home, decide that that’s not what she wants, and come up with a way to foil my intentions. She can plan and decide not to respond when I call her, and she understands that she has to be very still so I don’t see or hear her moving.

Cali hides deep in some thick bushes

She has to get the timing right, too. She wants to avoid going home, but it would be self-defeating to hide while I was still willing to play ball with her. So she has to read my body language carefully, in addition to parsing the verbal cues I give her. Finally, she also has to know when to resurface so as not to be abandoned (this would not happen, but how can she be sure of that?) or really get me angry.

A few days ago, she hid at the dog park. I might’ve very foolishly muttered something like “we need to get going” before throwing her ball. She took the ball, ducked under a branch, and disappeared. I called her. Nothing. I waited for about 10 minutes, then called some more. Nothing. I was pissed off. I had seen where she entered the bushes, and I went in after her. She was happy as can be, under some bushes with her ball. I grabbed the ball. That got her attention. We walked back to the car with me giving her an earful about what a terrible dog she was, bad to the bone. She laughed.

Then, the next day, she hid in a friend’s garden. I left her there (with the friend) and figured she’d have a nice afternoon in the grass, playing with dog and human friends. She did, but my friend reported that Cali cried when she realized that I had left. She recovered pretty quickly, though, and had fun; my friend also reported that Cali had a fence-fight with the next-door-neighbor-dog — which Cali instigated. But that’s a problem for another blog post.

I’m not really sure what to do about this new trick. I could just not take her to the dog park, but that’s a terrible solution. I don’t think that telling her to “come out right now or we won’t come back tomorrow” would work. She’s smart, but that is probably too abstract and verbose for her. I could punish her when I find her, but that would only motivate her to do a better job of hiding. I could keep her on a long leash, and risk getting it tangled in trees, picnic tables, other dogs …

She hides with her beloved tennis ball, and when she has her ball, she’s not interested in treats, so stooping to bribery is out. I could just wait her out, but I am not sure that my laptop battery would outlast her, and I do need to work.

Deni suggested letting Cali keep her tennis ball with her when we leave the park so that she’s not worried about losing it. In the apartment, all the time. Ick. Cali loves her tennis ball so much. She is happiest when it is really dirty, so she gets is all drooly and wet, then rolls it in dirt. Sometimes she digs a special pit just to make sure it gets totally covered in really fresh, dirty dirt. I’ve generally not allowed her to have the ball inside. If I get desperate enough, maybe …

Anyone have a better idea?