Sunday’s Missoulian (the local paper when I am in Montana) front page story, “Canine Calculator,” naturally caught my eye. Beau, a 12-year-old-Labrador, reportedly does math. Not only arithmetic, but algebra too. He also can figure out golf scores. Football, too. Check out the story and video if you don’t believe me. Actually, it was when I got to the claim about the algebra that I smiled at the idea that Missoula has its very own Clever Hans.
Clever Hans was a horse who, supposedly, could count. He lived in Germany in the early twentieth century. Hans was tested and performed his counting feats in front of many audiences. Ultimately, though, in 1907, Hans was defrocked. Controlled study of Hans found that all of his human questioners were unconsciously signaling Hans to stop counting (he tapped his hoof; Beau barks) at the correct number. Since then, the attribution of impossible feats to “clever” animals when, really, they are playing us for fools has been called the Clever Hans Effect.
Back to Beau and dogs. I have no doubt that dogs can count. I have no doubt that they can learn complex concepts and apply them. In fact, I have absolutely no doubt that dogs excel in physics and geometry — they can calculate exactly how far to run, when to jump, and how high to jump to snatch a Frisbee out of the air and make a graceful arc in landing. Even when it is a windy day on a Montana hilltop. They can watch a stick floating down a river and figure out how far and how fast to run and when to jump in to catch the stick. Few of us can do that, instantly, in our heads, as we’re running to catch the Frisbee or stick.
But dogs cannot calculate golf scores. And they cannot solve algebra problems.
They can, however, and do (every minute of our shared lives) watch us intently for cues we are not even aware we’re giving them. They notice and glean information from even the most subtle body language, as I learned when, years ago, trying to teach a puppy to sit, I inadvertently and completely unconsciously dipped my head every time I said, “sit.” It was obvious on video. It was even more obvious to the dog, who had decided that was part of the command, long before I realized I was doing it.
I’m much more careful now. But we’re often not aware of these almost imperceptible signals and might not even notice them on a video. The dogs (and horses) do. Everyone “testing” Hans displayed shifts in posture and facial expression as Hans reached the correct number of taps. The very subtle release of tension told Hans to stop counting. Rather than attributing “phenomenal” (and very human) skills to our animals, let’s appreciate the unique skills they have — and we lack.
Beau’s owner, on the Missoulian video, is doing the same thing as Hans’s testers. He says he’s not signaling Beau, and he’s probably not aware of it. But he is signaling his dog. He watches the dog intently as the dog barks. When Beau reaches the right number, the human’s eyes shift. As soon as the owner moves his eyes, Beau’s ears go forward and Beau stops barking. Beau has learned to look for and respond to that tiny cue.
Beau is said to respond with amazing accuracy, no matter who asks him math questions. I’d bet that nearly everyone gives him a similar signal — a relaxation of tension or shift of attention — and he’s expert at detecting those signals.
He’s not perfect, the owner acknowledges. In fact, on the video, Beau barks an incorrect answer. Why? The guy is talking to the reporter and glances up at the reporter. That instant, Beau stops barking. When the question is repeated and the owner’s attention stays focused on Beau, Beau gets it right.
So, you might be asking, what about those dogs who supposedly read? Is that also a Clever Hans Effect?
Jana reads. My longtime training mentor wrote an entire book about dogs reading, Teach Your Dog to Read. So, obviously, I think not. What’s the difference?
Dogs do learn to read words or pictures from flashcards and to perform the associated commands. The dog isn’t exactly reading as a form of entertainment, as we might. She is “reading” information from the flash cards, though. In fact, the reading dog is putting the same ability to work — detecting subtle visual cues and pairing meaning with them — as Beau.
The first word Jana read was “sit.” In Hebrew.
Wow, pretty amazing, right?
She simply learned to associate the picture or image — what looks like a Hebrew word to me — with sitting. She learned to “read” several English and Hebrew “words” and a few stick-figure pictures.
Once they get the concept of pairing an image with a particular action, dogs can learn new “words” or pictures very easily. Some dogs even grasp the concept that they should imitate what the stick-figure dog on the flash card is doing. Exceptional dogs are able, the first time they see the cards, to distinguish between a stick figure dog sitting and a stick figure dog sitting with one tiny, thin, stick-figure paw raised in a “shake.”
Just as Beau is “reading” the people for his cue to stop barking, Jana and her reading buddies are “reading” the images on the cards for a cue as to what behavior we crazy humans want them to do.
Why not? It’s easy for them to do, and it pays off in bountiful attention and many, many treats. Not a bad deal for those clever, calculating canines!
8 thoughts on “Canine Calculator Story Doesn’t Add Up”
[…] boards with which to network them together, are not the first to try to teach dogs to talk (or do math). Researcher and author Sean Senechal devised a dog sign language system several years […]
[…] many hoaxes and dubious claims, dogs can’t actually count, at least not without extensive training — but dogs are […]
Have you seen his video on missoulian.com?
Yes, several times. I also link to it and discuss it in the blog post. Have you read the blog post?
Thanks for your comment. I am not at all closed minded about dogs and their incredible intelligence. I write extensively about dogs’ problem solving abilities and even their moral development. I also believe that dog and human relationships work best when we appreciate each other for who we are — and that means acknowledging both our strengths and our weaknesses. I delight in dogs’ abilities to discriminate scent in ways that humans never could; I am fascinated by bats’ and dolphins’ abilities to use sonar and echolocation. Pointing out that humans cannot do those things is not demeaning to humans and does not take away from my love for and appreciation of the many human talents any more than saying that dogs can’t calculate golf scores detracts from my appreciation for dogs’ abilities. I would love to get to know Beau, as I am sure that is is a bright, talented dog. And if you can convince me that he can calculate golf scores, I will happily write an article admitting my error!
I think it is sad that you are to closed minded to believe that Beau can actually do math. The fact is that labs are extremely intelligent animals have proven their ability to be great problem solvers. If you truly do not believe in this dogs ability than contact me and we can make arrangements for you to spend some time with him one on one. I look forward to reading your recant 🙂
Great story! I didn’t watch the whole video, but from skimming the original story, it looks like the reporter didn’t consult any canine experts to explain Beau’s abilities.
I loved your explanation of how dogs “read”. It’s the same way that someone who is illiterate can get by in today’s world: associating images with meaning without deciphering their components.
But I’m unsure what you mean by this passage:
“[Dogs] can watch a stick floating down a river and figure out how far and how fast to run and when to jump in to catch the stick. We can’t do that, instantly, in our heads, as we’re running to catch the Frisbee or stick.”
Are you saying dogs can calculate trajectories and the like and humans can’t?
Well… I can’t, but I suppose some people can. Athletes. The ones who are really good at it get paid a ton of money!!
Maybe I should change that. We seem to need more tools to do stuff like that, though.