If A equals B, and B equals C, then it follows logically that A equals C, and even that C equals A. Most humans know that, sort of. So does Rio.
Rio, a seven-year-old California sea lion living at the University of California at Santa Cruz, thinks like a human. After learning to pair the silhouette of a crab with that of a tulip, and then the tulip with a radio, Rio makes a logical leap: she matches the crab with the radio. It may seem a small thing, but it is a basic kind of logical thinking. And Rio is the first nonhuman animal that is known to display it.
What’s more, says Rio’s trainer, animal behaviorist Ronald Schusterman of UCSC and California State University at Hayward, Rio’s ability undermines the common notion that language is a prerequisite for logical reasoning. Vocabulary, in this view, allows us to label and classify objects and to find equivalences among them--to recognize that two different animals are both crabs, say, and even to recognize that crab equals tulip in certain circumstances. But Rio does that without language, which suggests things may be the other way around: elementary logic may be a prerequisite, and an evolutionary precursor, of language.
I’m saying that forming equivalence relations is primitive, a basic, says Schusterman. Imagine if you took the specialized areas of the brain that do language and stuck them onto a cockroach. Do you think the cockroach would do language? Probably not. Because you need the precursors, a certain amount of intelligence to do language.
It was a desire to get beyond the inconclusiveness of animal language experiments that pushed Schusterman into his current line of research. Since the 1960s researchers have taught sign language to various animals, including chimpanzees, gorillas, orangutans, and dolphins. (Schusterman himself taught sea lions a language ten years ago.) In one experiment, a dolphin was taught to tap a ball to its right when given the signal right-ball-flipper-touch; in another, a chimp signed want berry when she had a hankering for the fruit.
But the significance of such experiments remains controversial. Some researchers believe the animals’ ability to learn the gestures or symbols, combine them into sequences, and use them in different contexts shows that the animals understand the meaning of the symbols and are displaying elementary reasoning. Other researchers, though, dismiss the animal-language results as sophisticated mimicry.
Schusterman decided to test Rio for reasoning ability in a different way. He and his graduate students created a repertoire of 90 objects for Rio to learn. Some were three-dimensional objects, such as a spatula, but most were simply black-on-white drawings. The researchers divided the objects into three categories (A, B, and C) and grouped them in 30 trios consisting of one object from each category.
As Rio sat in front of a windowed plywood panel at poolside, researchers behind the panel displayed an object from category A in the center window for a few seconds. Next they opened the two side windows, displaying the objects from category B that Rio had to choose between. If Rio touched her nose to the correct one, the researchers rewarded her with a herring.
Schusterman found that Rio readily learned to pair objects. For example, she learned that the silhouette of an elephant (category A) always went with that of an ant (category B). And she soon understood that the pairing was symmetric: if shown the ant first, she matched it with the elephant. Later, Rio learned the ant also matched the silhouette of the planet Saturn, an image from category C.
The crucial test came when Rio had to match the elephant (A) with Saturn (C)--images she had never seen together. In order to be able to do what we call equivalence--that is, to go from C to A immediately, without any special training--you would have to have an understanding of symmetry and transitivity, says Schusterman. That is, you’d have to know that once A equals B and B equals C, then A equals C. You’d also have to know that C equals A. Rio apparently knew all that: on her very first try, she made the correct A-C connections for 11 out of 12 image trios and the correct C- A connections for 17 out of the remaining 18 trios.
No other animal has yet matched Rio’s performance, according to Schusterman. But that doesn’t mean sea lions are smarter than dolphins or chimpanzees. I’m convinced that many animals have this ability, says Schusterman. They just haven’t been tested along these lines.