Dog who knew over 1,000 words dies

The New York Times reports:

John W. Pilley, a professor emeritus of psychology at Wofford College, taught his Border collie to understand more than 1,000 nouns. . . .

For three years, Dr. Pilley trained her four to five hours a day: He showed her an object, said its name up to 40 times, then hid it and asked her to find it. He used 800 cloth animal toys, 116 balls, 26 Frisbees and an assortment of plastic items to ultimately teach Chaser 1,022 nouns.

Chaser died on Tuesday at 15. She had been living with Dr. Pilley’s wife, Sally, and their daughter Robin in Spartanburg. Dr. Pilley died last year at 89. . . .

“What we would really like people to understand about Chaser is that she is not unique,” [John Bianchi’s daughter, Pilley Bianchi, who helped him train Chaser,] said. “It’s the way she was taught that is unique. We believed that my father tapped into something that was very simple: He taught Chaser a concept which he believed worked infinitely greater than learning a hundred behaviors.”

Ms. Bianchi said that her father’s experiment was “uncharted territory” in animal cognition research, pointing to news media coverage calling Chaser “the world’s smartest dog.”

“Her language learning is very high-level, powerful science,” she said. . . .

If Chaser had 30 balls, Ms. Bianchi said, she would be able to understand each one by its proper-noun name and also as a part of a group of objects. “She learned the theory of one to many and many to one, which is learning one object could have many names and many names can apply to one object or one person,” she said.

Greg Nelson, a veterinarian at Central Veterinary Associates in Valley Stream, N.Y., said humans were learning that animals have a deeper understanding of the world around them.

“People have always been under the belief that animals respond to commands based on a rewards system,” he said. “Learn limited commands and tricks, then get a treat.”

But “they do have a language among themselves, spoken and unspoken,” he added. “And it’s apparent that they can understand the human language probably in much the same way as we learn a foreign language.”

You can say the communication in this video starting at 2:08. At first I thought she could be responding to Dr. Pilley's nonverbal cues, as when the owner would move in the direction of the frisbee while asking her where the frisbee is. But then she really does seem to be understanding language when he says: "Chase, to Powderpuff [a doll's name], take frisbee."



Dr. Pilley says in the video:

These kinds of findings definitely show that lower animals, especially dogs, are not just machines with blood. They have emotions, they have mental processes.

There's a better demonstration here, as a seemingly skeptical Neil deGrasse Tyson asks Chaser to find certain toys that are all out of Tyson's view (after 2 minutes in). "I asked Chaser to find 9 toys, and she got every one right. And . . . I chose the toys from this huge pile; neither John [Pilley] nor Chaser saw which ones I picked." She also made a logical inference: when Tyson asked her to find a doll she had never heard of before, "Darwin," out of a group of 9 toys, she chose the only toy she had never seen before.



So I think those videos prove the dog really did understand the words. In the past, I've been willing to call BS on claims of animals with supposedly sophisticated language understanding that seem like scams, as I did with Koko the gorilla (see my last comment in this Facebook post, quoting a skeptical Slate article).

With these kinds of claims of an animal understanding language, there are always going to be those who question whether the animal really has that linguistic understanding, or if what's really going on is the animal is picking up on other cues from the owner. So it's important to take that skepticism seriously and address it directly, in order to show people how much the animal understands.