Partway into considering examples of dolphin “cognition,” I realized that dolphins are as cognizant as anyone. There are so many examples of them acting aware and clever (because they are aware and clever) that one might as well try to compile examples of humans acting aware and clever. That’s just who we are. And it’s who dolphins are, too. Dolphins and humans have not shared a common ancestor for tens of millions of years. Yet for all the seeming estrangement of lives lived in liquid, when they see us they often come to play, and we greet them and can recognize in those eyes that someone very special is home. “There is someone in there. It’s not a human, but it is a someone,” says dolphin expert Diana Reiss.


Academic study of dolphin intelligence got off to a rickety start that cost it about a decade. In some sense it has never recovered from its first publicly noted researcher, who cloaked dolphins with a mystical allure they’ve never quite shaken. On the other hand, dolphins have earned a bit of mystique.

In the late 1950s and ’60s, neurophysiologist and brain researcher John C. Lilly presented us with creatures whose gigantic brains made them our superiors. It was an improvement on the idea that whales felt nothing but an inexplicable urge to swallow humans. But Lilly, too, was wrong. John Lilly pronounced that an animal with a brain the size of a sperm whale’s must have a “truly godlike” mind. We’ll leave aside the question of what a “godlike” mind would be and what a whale would do with one. Lilly mistakenly assumed that brain size translated directly to thinking ability.

Different species’ brains emphasize different abilities. The nerves and brain structure for detecting and analyzing scent are important parts of a dog’s brain but essentially nonexistent in a whale’s. Meanwhile, a sperm whale’s brain devotes enormous resources to creating, detecting, and analyzing sound. Sperm whales’ brains are larger than blue whales’, though blue whales’ bodies are twice as big. What does a sperm whale do with its singular brain? It sets courses for long migrations and keeps track of family and friends over decades and across thousands of miles of travel. It prepares for dives deeper than a mile; manages the pumping, distribution, and shunting of blood and oxygen while the whale stops breathing for up to two hours; and controls the tracking and muscle coordination needed while hunting squid the size of your nightmares in total darkness. It does some things that humans cannot do, and it cannot do some things that humans do. It’s much more interesting and much more useful for the tasks at hand than anything “truly godlike.” “Godlike,” anyway, was a grandiose Band-Aid for “We don’t know.” It covered a big intellectual boo-boo in Lilly’s own thinking.

Scientists rightly scorned John Lilly. His insistence that we could crack dolphin communication — by teaching them English — proved wrong. But his image of dolphins as superior to humans grabbed the imagination of the public, which remains captivated, waiting for a sign that they’re on a higher plane. Perhaps we hope that somehow, someday, someone better will deliver us from our own evils.

It wasn’t until the 1970s and the work of Louis Herman’s group that inquiry into dolphin cognition got real. Herman showed that a Hawaiian bottlenose dolphin named Akeakamai could respond correctly when shown an arbitrary symbol (not a literal picture) for “ball,” followed by a symbol for “question.” If there was no ball, she would press a “No” lever. This demonstrated that a dolphin could form a concept of a ball, and that she could call upon that knowledge when presented with a symbol used to represent “ball.” It showed that dolphins are, as long suspected, very intelligent. Whatever “intelligent” means.


While dolphins have grasped a bit about human language in the pools where they learned the signs and symbols used by researchers, we’ve never cracked their code or figured out how to use dolphin sounds to communicate with dolphins about dolphin things. Do they talk to each other and give each other commands and instructions and tell stories? We don’t know. What they’re thinking: we don’t know. What they’re saying: we don’t know. Can we begin to know?

Like human babies, infant dolphins babble sequences of whistles that become more organized as they grow. At anywhere between one month and two years, bottlenose, Atlantic spotted, and other dolphins develop their own distinctive individual “signature whistles.” Signature whistles are a name they create for themselves. The sound is distinctive, and the dolphin doesn’t change it, ever. They use it to announce themselves.

Dolphins who hear their own signature whistled by another dolphin call back. They don’t respond to a dolphin who whistles a third dolphin’s signature call. In other words, they call each other by name, and they answer when they hear their own name called. Dolphins call their close friends’ names when they are separated. No other mammal seems to do that (that we know of). Dolphins more than ten miles away can hear each other if water conditions are right. Atlantic spotted dolphins seemingly use names to call together several individuals. When groups meet at sea, they exchange names (but not, so far as is known, phone numbers).


Dolphins remember and recognize one another’s signature whistles for their whole lives. In the experiment that showed this, captive bottlenose dolphins heard recorded signature whistles of dolphins with whom they’d been housed as long as twenty years earlier. They remembered and responded even if they’d known each other only a short while before being separated. The experimenter, Jason Bruck, concluded, “Dolphins have the potential for lifelong memory for each other.” That was the first formal study showing social memory lasting twenty years in a nonhuman.


Dolphins at the Institute for Marine Mammal Studies, in Mississippi, were trained to help keep their pools clean by trading litter for fish. A dolphin named Kelly realized that she got the same size fish for bringing a big sheet of paper as for a small piece. So, under a weight at the bottom of the pool, she hid any paper that blew in. When a trainer passed, she tore off a piece of paper to trade for a fish. Then she tore off another piece, got another fish. Into the economy of litter, she’d rigged a kind of trash inflation rate that kept the food coming. Similarly, in California, a dolphin named Spock got busted for tearing pieces off a paper bag he’d stuffed behind one of the pool’s underwater pipes, using each shred to buy another fish.

One day, a gull flew into Kelly’s pool, and she grabbed it and waited for the trainers. The humans seemed to really like birds; they traded her several fish for it. This gave Kelly a new insight, and a plan: During her next meal she took the last fish and hid it. When the humans left, she brought the fish up and baited more gulls, to get even more fish. After all, why wait to scrounge an occasional piece of accidental paper when you could become a wealthy commercial bird-fishing dolphin? She taught this to her youngster, who taught other youngsters, and so the dolphins there become professional gull baiters.

A young killer whale at Marineland Canada, in Ontario, somehow figured out that spreading mashed fish at the surface of the pool, then sinking out of view, could bring a little sport into his life. If a gull landed, the whale shot upward, sometimes catching — and eating — the gull. He set the trap many times. Eventually his younger half brother and three other whales caught on.

Insight, innovation, planning, culture.

In 1979 Diana Reiss started working with a captive bottlenose dolphin named Circe. When Circe did the behavior that Reiss was looking for, Circe got verbal praise and some fish. When she didn’t, she got a “time-out,” in which Reiss stepped back or turned away to indicate that Circe had not performed “correctly.” (Time-outs are now considered outdated; they can frustrate intelligent creatures.) Circe didn’t like tail fins left on her mackerel, and by spitting out the pieces with tails, she essentially trained Reiss to cut them off. One day a few weeks into training, Reiss absentmindedly gave Circe an untrimmed tail section. Circe waved her head from side to side the way we might indicate “No,” spat out the fish, swam to the other side of the pool, positioned herself upright, and just looked at Reiss for a short time. Then she came back. Circe the dolphin had given Reiss the human a time-out.

Astonished but skeptical, Reiss planned an experiment. Six times over several weeks, Reiss purposely fed Circe a tail section with the fin on. Circe gave Reiss four more time-outs. Those were the only times Circe behaved that way. Circe had not only learned “reward” and “no reward; time-out” for her own behavior; she had conceptualized the time-out as a way of communicating the idea “That’s not what I’ve asked for” and used it to correct her human friend.

Reiss also worked with a young male named Pan. Pan was learning to use abstract symbols on a keypad. (The symbols were never literal; the symbol for “ball” might be a triangle. And the keys were moved around, so that the dolphins had to learn the symbol, not the location, for what they wanted.) Pan didn’t care about toys; he really wanted fish. When Reiss removed the fish key from the options, Pan found a fish left over from the morning’s breakfast, swam to the keyboard, touched the fish to a blank key, and expectantly looked Reiss in the eye. Reiss understood exactly what he wanted; Pan was making himself very clear.

Not long into the project, the dolphins started copying the different whistle sounds that the computer paired with the various objects. When Pan and his poolmate, Delphi, played with the toys, they copied the computer sounds for “ball,” “ring,” and other objects. Reiss explained this to me, then added, “One day I had given Pan a fetch signal. There was only one toy in the pool, a ball, but Delphi had it in her mouth. Pan swam over to Delphi, and I heard someone do the ‘ball’ whistle. Delphi passed it to Pan, and they both swam to me with it.” They’d learned human symbols and used them for communication with each other.

Another dolphin, also named Delphi but male, started to play with his food, holding fish in his mouth and then dropping them all over the pool. Reiss trained Delphi to understand the command “swallow” and would not give him another fish until he had shown that the first was indeed gone. This worked for the next week, while Reiss was away; her students fed Delphi and demanded proof of his “swallow.” When Reiss came back, Delphi’s swallows looked exaggerated. Sore throat? More exaggerated gulps, more displays of his empty mouth, more fish given. Suddenly, writes Reiss, “Delphi’s eyes got really big.” He opened his mouth. Seafood; see — food! “I saw all these whole fish in there.” He must have been holding them in his throat. “Before I had time to open my mouth in surprise, he started to shake his head, left and right, left and right.” Fish flew everywhere. “Delphi was obviously having fun, and he had chosen to play this trick on me, not one of the students.” Delphi had utterly fooled and manipulated Reiss — and he seemed to relish it. As did Reiss, who says, “I laughed hysterically.”

Excerpted from Beyond Words: What Animals Think and Feel, by Carl Safina. Copyright © 2015 by Carl Safina. Reprinted by permission of Henry Holt and Company.