Can Dogs Talk Using Buttons?

[BillTre]

Here is a nicely done PBS documentary on the ability of dogs to communicate.
There are several different approaches to possible dog communication, but its mostly focused on dogs pushing talk buttons. These are mostly on home button boards that make speech sounds when the dog pushes it (like "outside"). Since they become popular, many are using boards that can be linked up to a data collection site. They have thousands of dogs and millions of button pushes.
There are also several anecdotes indicative of understanding of the words that were button pushed.

Dogs are probably the animal most closely associated with people for the longest time (40,000 years?). Probably followed by cats. Among the domesticated animals, horses probably have the longest and most intense history of interactions (since horse riding) vs. goats, sheep, etc.).

 

[256bits]

Video is unavailable as a general Youtube outside of the USA.
One needs a PBS account, VPN, or some other mystical technological override.

Nonetheless, looking it up, button talking dogs seem to be all the rage for some owners to show off how 'smart' their dog is. Their dog is their 'child'?
A fun project for the pet owner.

Animals pick up easily on subtle visual gestures.
And gestures that most humans consciously miss.
See https://en.wikipedia.org/wiki/Clever_Hans

My impression is similar to this lady biologist.
There is something there.
But the spoken 'language' interpretation that some owners like to give credit to their favourite pet seems to be extensively overplayed.

 

[Rive]

The linked video is not accessible for me, either...

But the spoken 'language' interpretation that some owners like to give credit to their favourite pet seems to be extensively overplayed.

Well... On one side, by now we know that communication issues and mismatched conceptual systems makes evaluations of intelligence inherently problematic.

On other side, AI just show us that actually how overplayed was the self-evaluation of our own communication.

I think within only a few (dog) generations there will be some geniuses with some kind of (measured) intelligence comparable to children (of various ages). And then on it'll be about numbers.

 

[BillTre]

Its too bad you could not see the actual video. The pet owner aspect is dealt with pretty well in the video.
It is a complex examination of the subject that a brief post will not do justice to.
The question of the dogs' understanding is examined in several ways, both statistically (based on the huge number of pooled responses and anecdotes and followups by some scientists.
Understanding by the dogs could be explained away in many but not all cases. A small number dogs were extraordinary in their abilities. I have had some rather dumb dogs, but also some smarter ones, so I think variance exists in dog smarts.
On the other hand, its understanding of single or pairs of words not full sentences like people.
Some of their little experiments eliminated gestures and other clues.

WRT the 2nd video:
She makes some good points, but several were addressed in the video.
Dogs have had something like 20,000 years to evolve improved communication with people, so some improvements over non-domesticated critters would be expected. They have evolved many other more easily observable changes.
The first video is not a conclusive exploration of the subject, but it does address several of the issues in the second video.

 

[berkeman]

There are several different approaches to possible dog communication, but its mostly focused on dogs pushing talk buttons. These are mostly on home button boards that make speech sounds when the dog pushes it (like "outside"). Since they become popular, many are using boards that can be linked up to a data collection site. They have thousands of dogs and millions of button pushes.
There are also several anecdotes indicative of understanding of the words that were button pushed.

I keep seeing color coded buttons or button groups in your video and other photos. Why do they use color coding since dogs are red-green color blind? I guess they can see yellow-blue differences, but orange-pink-green?

 

[BillTre]

Good catch.
I would guess the colors are for the humans who buy them.
The dogs should only perceive differences in intensity of light (perhaps due to saturation differences).
It raises the question: are the different colors always paired with the same colors.

 

[symbolipoint]

I think within only a few (dog) generations there will be some geniuses with some kind of (measured) intelligence comparable to children (of various ages). And then on it'll be about numbers.

Those already exist and are commonly understood to be the working breeds, such as Doberman, etc...

 

[symbolipoint]

Regarding post #2, I will agree with KP, from the video.

 

[DaveE]

One of the big problems I have as an instructor of sport detection dogs is getting the humans to recognize that dogs neither think like humans or perceive the world as humans do. Now, anytime people talk about animal intelligence, I'll pretty much check out. It always comes from the frame of human thought, human values, the human umwelt. Cockroaches are really smart at being cockroaches, etc.

If you really work with and care about dogs AS DOGS, this is obvious. Most people think of dogs as an extension of their human world. I can spot a "dog person" a mile away, most aren't. This is why when I'm walking my dog on the suburban sidewalks or at the park, we avoid other dogs. It's not really about the other dog, it's all about the probably clueless dog owner.

How about learning to speak dog, instead of teaching them to speak human? It can be done... if you care to.

BTW, one thing you won't find on PF is animal behavior people.

 

[symbolipoint]

DaveE said this part:

How about learning to speak dog, instead of teaching them to speak human? It can be done... if you care to.

Yes. We should. But we just wish and hope the dogs will teach this to us (or maybe some of them are trying but we are not paying careful enough attention).

DaveE said this, too:

BTW, one thing you won't find on PF is animal behavior people.

Why? That needs to change.

Much like what you said about cockroaches, dogs are excellent at being dogs, but you'll need to check with them about how well they communicate with each other.

 

[Rive]

Those already exist and are commonly understood to be the working breeds, such as Doberman, etc...

That's the obscure intelligence of those breeds within their own area of expertise. For common people it's not trivial to be related with it.
What I'm talking about is, that there will be a new breed (direction of selection) emerging with expertise in (buttoned) human relations.

 

[weirdoguy]

it's all about the probably clueless dog owner.

Oh I hate most dog owners. Especially those whose dogs run around without leash and approach my dog. My dog also always walks without leash, but he does not run to other dogs, he sticks with me. And my favourite "He/she won't do anything!"... Well, "But I will." Anthony has been attacked couple of times, and he's alive only because he has thick fur.

 

[symbolipoint]

That's the obscure intelligence of those breeds within their own area of expertise. For common people it's not trivial to be related with it.
What I'm talking about is, that there will be a new breed (direction of selection) emerging with expertise in (buttoned) human relations.

You are maybe suggesting the idea of selective breeding to promote this kind of communication between Dog and Human. One must wonder about the ethics of that kind of breeding.

Excuse me while I "Insert quotes..." because i accidently clicked on Multiquote button.
and I cannot get rid of the quote frame after removing what was inside the quote frame.

 

[Rive]

You are maybe suggesting the idea of selective breeding to promote this kind of communication between Dog and Human.

I'm not exactly suggesting the idea. It'll just happen.
...actually, it's not that bad of a career for a 'city dog'. Degradation of 'working breeds' in favor of primarily being pets/company is a long established process.

 

[DaveE]

What I'm talking about is, that there will be a new breed (direction of selection) emerging with expertise in (buttoned) human relations.

It will be hard to select more expertise in human relations, since they are already experts. The pinnacle of domestication, they see more about our behavior than we know. Dogs know how to communicate with humans for the important stuff, but do the humans care to listen? Many obnoxious dog behavior issues are the result of the dog knowing "you can't ignore me if I do this".

This whole button pushing stuff just looks like an extension (quantitively, not qualitatively) of simple associative learning. If I do this behavior, I get this result. It's not language. It's not different than the Border Collie that knows the names of 1000 toys*. It's not different than our dogs when they want to come in the house through the back door. One barks softly exactly once and waits, the other gently scratches the glass exactly once and waits. They taught me this language, they chose it. Many, many dogs have been trained to poke bells hanging from the door knob. This could easily be trained with a colored button with a label "open the door"; that's not any smarter than barking a certain way.

Also, every video I've seen about this subject is a monument to awful research. No blinding, no controls, etc. It's 99% humans making up stories about what their dog knows, usually post hoc. There is some real research on this, but you won't find it in a YouTube video.

* As an aside: The interesting part of the studies with Chaser wasn't the number, if your dog knows the name of three toys teaching a forth is simple. It was the experiment where they put out a bunch of toys he knew plus one new one. They asked him to get the new one with a name he'd never heard before. He quickly went and got it. So, a bit of inferential learning.

 

[berkeman]

Degradation of 'working breeds' in favor of primarily being pets/company is a long established process.

 

[DaveE]

This discussion reminded me of this experiment. Could it be that humans are as smart as dogs? Look they can track things with their nose!