What Did You Say?
Head trainer Teri Turner Bolton looks out at two young adult male dolphins, Hector and Han, whose beaks, or rostra, are poking above the water as they eagerly await a command. The bottlenose dolphins at the Roatán Institute for Marine Sciences (RIMS), a resort and research institution on an island off the coast of Honduras, are old pros at dolphin performance art. They’ve been trained to corkscrew through the air on command, skate backward across the surface of the water while standing upright on their tails, and wave their pectoral fins at the tourists who arrive several times a week on cruise ships.
But the scientists at RIMS are more interested in how the dolphins think than in what they can do. When given the hand signal to “innovate,” Hector and Han know to dip below the surface and blow a bubble, or vault out of the water, or dive down to the ocean floor, or perform any of the dozen or so other maneuvers in their repertoire—but not to repeat anything they’ve already done during that session. Incredibly, they usually understand that they’re supposed to keep trying some new behavior each session.
Bolton presses her palms together over her head, the signal to innovate, and then puts her fists together, the sign for “tandem.” With those two gestures, she has instructed the dolphins to show her a behavior she hasn’t seen during this session and to do it in unison.
Hector and Han disappear beneath the surface. With them is a comparative psychologist named Stan Kuczaj, wearing a wet suit and snorkel gear and carrying a large underwater video camera with hydrophones. He records several seconds of audible chirping between Hector and Han, then his camera captures them both slowly rolling over in unison and flapping their tails three times simultaneously.
Above the surface Bolton presses her thumbs and middle fingers together, telling the dolphins to keep up this cooperative innovation. And they do. The 400-pound animals sink down, exchange a few more high-pitched whistles, and then simultaneously blow bubbles together. Then they pirouette side by side. Then they tail walk. After eight nearly perfectly synchronized sequences, the session ends.
There are two possible explanations of this remarkable behavior. Either one dolphin is mimicking the other so quickly and precisely that the apparent coordination is only an illusion. Or it’s not an illusion at all: When they whistle back and forth beneath the surface, they’re literally discussing a plan.
When a chimpanzee gazes at a piece of fruit or a silverback gorilla beats his chest to warn off an approaching male, it’s hard not to see a bit of ourselves in those behaviors and even to imagine what the animals might be thinking. We are, after all, great apes like them, and their intelligence often feels like a diminished—or at least a familiar—version of our own. But dolphins are something truly different. They “see” with sonar and do so with such phenomenal precision that they can tell from a hundred feet away whether an object is made of metal, plastic, or wood. They can even eavesdrop on the echolocating clicks of other dolphins to figure out what they’re looking at. Unlike primates, they don’t breathe automatically, and they seem to sleep with only half their brains resting at a time. Their eyes operate independently of each other. They’re a kind of alien intelligence sharing our planet—watching them may be the closest we’ll come to encountering ET.
Dolphins are extraordinarily garrulous. Not only do they whistle and click, but they also emit loud broadband packets of sound called burst pulses to discipline their young and chase away sharks. Scientists listening to all these sounds have long wondered what, if anything, they might mean. Surely such a large-brained, highly social creature wouldn’t waste all that energy babbling beneath the waves unless the vocalizations contained some sort of meaningful content. And yet despite a half century of study, nobody can say what the fundamental units of dolphin vocalization are or how those units get assembled.
“If we can find a pattern connecting vocalization to behavior, it’ll be a huge deal,” says Kuczaj, 64, who has published more scientific articles on dolphin cognition than almost anyone else in the field. He believes that his work with the synchronized dolphins at RIMS may prove to be a Rosetta stone that unlocks dolphin communication, though he adds, “The sophistication of dolphins that makes them so interesting also makes them really difficult to study.”
Yet virtually no evidence supports the existence of anything resembling a dolphin language, and some scientists express exasperation at the continued quixotic search. “There is also no evidence that dolphins cannot time travel, cannot bend spoons with their minds, and cannot shoot lasers out of their blowholes,” writes Justin Gregg, author of Are Dolphins Really Smart? The Mammal Behind the Myth. “The ever-present scientific caveat that ‘there is much we do not know’ has allowed dolphinese proponents to slip the idea of dolphin language in the back door.”
But where Gregg sees a half century of failure, Kuczaj and other prominent researchers see a preponderance of circumstantial evidence that leads them to believe that the problem simply hasn’t yet been looked at in the right way, with the right set of tools. It’s only within the past decade or so that high-frequency underwater audio recorders, like the one Kuczaj uses, have been able to capture the full spectrum of dolphin sounds, and only during the past couple of years that new data-mining algorithms have made possible a meaningful analysis of those recordings. Ultimately dolphin vocalization is either one of the greatest unsolved mysteries of science or one of its greatest blind alleys.
Until our upstart genus surpassed them, dolphins were probably the largest brained, and presumably the most intelligent, creatures on the planet. Pound for pound, relative to body size, their brains are still among the largest in the animal kingdom—and larger than those of chimpanzees. The last common ancestor of humans and chimps lived some six million years ago. By comparison cetaceans such as dolphins split off from the rest of the mammal lineage about 55 million years ago, and they and primates haven’t shared an ancestor for 95 million years.
This means that primates and cetaceans have been on two different evolutionary trajectories for a very long time, and the result is not only two different body types but also two different kinds of brains. Primates, for example, have large frontal lobes, which are responsible for executive decision-making and planning. Dolphins don’t have much in the way of frontal lobes, but they still have an impressive flair for solving problems and, apparently, a capacity to plan for the future. We primates process visual information in the back of our brains and language and auditory information in the temporal lobes, located on the brain’s flanks. Dolphins process visual and auditory information in different parts of the neocortex, and the paths that information takes to get into and out of the cortex are markedly different. Dolphins also have an extremely well developed and defined paralimbic system for processing emotions. One hypothesis is that it may be essential to the intimate social and emotional bonds that exist within dolphin communities.
“A dolphin alone is not really a dolphin,” says Lori Marino, a biopsychologist and executive director of the Kimmela Center for Animal Advocacy. “Being a dolphin means being embedded in a complex social network. Even more so than with humans.”
When dolphins are in trouble, they display a degree of cohesiveness rarely seen in other animal groups. If one becomes sick and heads toward shallow water, the entire group will sometimes follow, which can lead to mass strandings. It’s as if they have a singular focus on the stranded dolphin, Marino says, “and the only way to break that concentration may be to give them something equally strong to pull them away.” A mass stranding in Australia in 2013 was averted only when humans intervened, capturing a juvenile of the group and taking her out to the open ocean; her distress calls drew the group back to sea.
Why did dolphins, of all the creatures roaming land and sea, acquire such large brains? To answer that question, we must look at the fossil record. About 34 million years ago the ancestors of modern dolphins were large creatures with wolflike teeth. Around that time, it’s theorized, a period of significant oceanic cooling shifted food supplies and created a new ecological niche, which offered dolphins opportunities and changed how they hunted. Their brains became larger, and their terrifying teeth gave way to the smaller, peglike teeth that dolphins have today. Changes to inner-ear bones suggest that this period also marked the beginnings of echolocation, as some dolphins likely changed from solitary hunters of large fish to collective hunters of schools of smaller fish. Dolphins became more communicative, more social—and probably more intelligent.
Richard Connor, who studies the social lives of dolphins in Shark Bay, Australia, has identified three levels of alliances within their large, open social network. Males tend to form pairs and trios that aggressively court females and then keep those females under close guard. Some of these pairs and trios are remarkably stable relationships that can last for decades. Males are also members of larger teams of 4 to 14, which Connor dubs second-order alliances. These teams come together to steal females from other groups and defend their own females against attacks, and they can remain intact for 16 years. Connor has observed even larger, third-order alliances that coalesce when there are big battles between second-order alliances.
Two dolphins can be friends one day and foes the next, depending on which other dolphins are nearby. Primates tend to have a “you’re either with us or against us” mentality when it comes to making distinctions within and between groups. But for dolphins, alliances seem to be situational and extremely complicated. The need to keep track of all those relationships may help explain why dolphins possess such large brains.
Dolphins are also among the most cosmopolitan animals on the planet. Like humans on land, dolphin species are seemingly everywhere in the sea, and like humans, they have proved ingenious at discovering feeding strategies that are particular to the environments they inhabit. In Shark Bay some bottlenose dolphins detach sponges from the seafloor and place them on their beaks for protection while searching the sand for small hidden fish—a kind of primitive tool use. In the shallow waters of Florida Bay dolphins use their speed, which can exceed 20 miles an hour, to swim quick circles around schools of mullet fish, stirring up curtains of mud that force the fish to leap out of the water into the dolphins’ waiting mouths. Dusky dolphins off the coast of Patagonia herd schools of anchovies into neat spheres and then take turns gulping them down.
All these behaviors have the mark of intelligence. But what is intelligence really? When pressed, we often have to admit that we’re measuring how similar a species is to us. Kuczaj thinks that’s a mistake. “The question is not how smart are dolphins, but how are dolphins smart?”
There are people who go on spiritual retreats to commune with dolphins, women who choose to give birth in the presence of dolphins, and centers that claim to use the powers of dolphin energy to treat the sick. “There are probably more weird ideas about dolphins swimming in cyberspace than there are dolphins swimming in the ocean,” writes Gregg. Many of those weird ideas can be traced back to a single man, named John Lilly.
Lilly was an iconoclastic neurophysiologist at the U.S. National Institute of Mental Health who began studying dolphins in the 1950s. In best-selling books like Man and Dolphin: Adventures on a New Scientific Frontier and The Mind of the Dolphin: A Nonhuman Intelligence, he was the first scientist to posit that these “humans of the sea” had a language. Almost single-handedly, writes Gregg, he “managed to transform what was initially regarded as an odd air-breathing fish at the turn of the 20th century into an animal whose intelligence is so sophisticated that it deserves the same constitutional protection as you or me.”
With grants from major scientific funding bodies, Lilly opened a dolphin research facility in the U.S. Virgin Islands, where attempts were made to teach a dolphin named Peter to speak English. As the 1960s dawned, Lilly’s experiments grew more and more unconventional—at one point he injected dolphins with LSD—and his funding began to dry up. He wandered off into the weirdest corners of the counterculture and carried with him the credibility of the field he’d helped create. Dolphin “language” would become an untouchable subject until 1970, when a University of Hawaii psychologist named Louis Herman founded the Kewalo Basin Marine Mammal Laboratory in Honolulu.
“We wanted to educate them to reveal their cognitive potential,” says Adam Pack of the University of Hawaii at Hilo, who worked at the lab for 21 years. “We reared the dolphins as you would a child.”
At Kewalo Basin two captive bottlenose dolphins, Phoenix and Akeakamai, were raised in an environment of constant education and schooled in an artificial language. Both were taught to associate either sounds or hand signs with objects, actions, and modifiers.
But Phoenix was taught an acoustic language in which words were placed in the order of the tasks to be performed. Akeakamai was taught a gestural language in which the order of the words was not the same as the order of the tasks. Though Phoenix could in theory respond word by word, Akeakamai could interpret her instructions only after she’d seen the entire sequence of gestures. Swimming in a pool filled with objects, the dolphins would carry out their instructions correctly more than 80 percent of the time.
After Akeakamai died in 2003 and Phoenix in 2004, their ashes were taken out to sea on surfboards and scattered, and the only research facility in the world dedicated solely to understanding how dolphins think went out of business. A big question remained: Why had Phoenix and Akeakamai found it so easy to learn the languages? Herman dismisses any notion that the researchers were piggybacking on some innate linguistic capacity. In his view, the imposed languages had allowed Phoenix and Akeakamai to express exceptional cognitive abilities common to all bottlenose dolphins—and perhaps other dolphin species—in a way that might never be exhibited in the wild. But is there some native form of dolphin communication that humans could eavesdrop on and eventually understand?
It turns out that there’s strong evidence to suggest that at least one kind of dolphin sound, studied extensively over the past decade, does function as a kind of referential symbol. Dolphins use distinct “signature whistles” to identify and call to one another. Each dolphin is thought to invent a unique name for itself as a calf and to keep it for life. Dolphins greet one another at sea by exchanging signature whistles and seem to remember the signature whistles of other dolphins for decades. Though other species, like vervet monkeys and prairie dogs, make sounds that refer to predators, no other animal, besides humans, is believed to have specific labels for individuals.
Signature whistles are only some of the vocalizations dolphins make underwater. What are the chances that they’re the only sounds in the dolphins’ repertoire that refer to something? How likely is it that dolphins have names only for each other and not for anything else in the sea?
A veritable Jane Goodall of the sea, Denise Herzing has spent the past three decades getting to know more than 300 individual Atlantic spotted dolphins spanning three generations. She works a 175-square-mile swath of ocean off the Bahamas, in the longest running underwater wild-dolphin program in the world. Because of its crystal clear waters, it’s a place where dolphin researchers can spend extended periods observing and interacting with wild animals.
Last summer I joined Herzing aboard her research boat, the R.V. Stenella, as she was preparing to run her first live trials with a complex new piece of machinery that she hopes will someday enable two-way communication between herself and the dolphins she has spent so long getting to know—and along the way illuminate how they communicate among themselves.
That piece of machinery is a shoebox-size cube of aluminum and clear plastic known as CHAT (cetacean hearing and telemetry), which Herzing wears underwater strapped to her chest. The 20-pound box has a small speaker and keyboard on its face and two hydrophones that look like eyes sticking out below. Inside, amid a tangle of wires and circuit boards sealed off from the corrosive effects of seawater, is a computer that can broadcast dolphins’ prerecorded signature whistles as well as dolphin-like whistles into the ocean at the push of a button and record any sounds that dolphins whistle back. If a dolphin repeats one of the dolphin-like whistles, the computer can convert the sound into words and then play them through a headset in Herzing’s ear.
Dolphins are notoriously talented mimics and quick students. Herzing’s goal is to get a handful of juvenile females she has known since birth to associate each of three whistle sounds broadcast by the CHAT box with a specific object: a scarf, a rope, and a piece of sargassum, a brown seaweed that wild dolphins use as a toy. Those three “words,” she hopes, will form the rudiments of a growing vocabulary of whistles shared by her and her dolphins—the beginnings of an artificial language in which she and they might someday be able to communicate.
“Once they get it—like Helen Keller getting language—we think it’s going to go very rapidly,” Herzing says. “Because they’re social, we’re capitalizing on other individuals watching. It’s like kids on a playground.”
Herzing, 58, is buoyant and optimistic, the kind of person for whom the word “visionary”—with its implications of both genius and kookiness—seems fitting. When she was 12 years old, she entered a scholarship contest that required her to answer the question “What would you do for the world if you could do one thing?” Her reply: “I would develop a human-animal translator so that we can understand other minds on the planet.”
In her underwater sessions, face-to-face with dolphins, sometimes for hours at a time, Herzing has recorded and logged thousands of hours of footage of every kind of dolphin behavior. She has also assembled a huge database of her loquacious subjects’ vocalizations.
Aboard the Stenella was another notable scientist, Thad Starner, a professor of computing at Georgia Tech. A pioneer of wearable computers, he’s also a technical lead at Google, where he works on Glass, the heads-up display that allows wearers to access the Internet as they go about their day.
Starner, 45, is boyish, with curly blond hair, wide eyes, and bushy sideburns. He wears Glass pretty much all the time and takes notes with a lemon-shaped keyboard that’s strapped to his left hand and fits in his palm. Starner’s lab team fabricated the CHAT box, and he’s come aboard the boat for ten days of technical testing and data collection.
If the mysteries of dolphin communication are ever to be cracked, it may have less to do with the two-way CHAT boxes than with the data-analysis tools Starner and his students have begun applying to Herzing’s dolphin recordings. They’re designing an algorithm that systematically searches through heaps of uncategorized data to find the fundamental units hiding inside. Feed in videos of people using sign language, and the algorithm pulls the meaningful gestures out of the jumble of hand movements. Feed in audio of people reading off phone numbers, and it figures out that there are 11 fundamental digits. (It’s not smart enough to realize that “zero” and “O” are the same number.) The algorithm uncovers recurring motifs that might not be obvious and that a human might not know how to look for.
As an early test of the algorithm, Herzing sent Starner a set of vocalizations she’d recorded underwater without telling him that he was listening to signature whistles sent between mothers and calves. The algorithm pulled five fundamental units from the data, which suggested that signature whistles were made up of individual components that were repeated and consistent between mothers and calves and that might be recombined in interesting ways.
“At some point we want to have a CHAT box with all the fundamental units of dolphin sound in it,” says Starner. “The box will translate whatever the system is hearing into a string of symbols and allow Denise to send back some string of fundamental units. Can we discover the fundamental units? Can we allow her to reproduce the fundamental units? Can we do it all on the fly? That’s the holy grail.”
When the opportunity finally arrives to test the CHAT box in the wild, it’s not just any dolphins that show up at the bow of the Stenella. The two dolphins that swim up to the boat are ones that Herzing has been hoping to encounter all week: Meridian and Nereide. Indeed, recordings of both dolphins’ signature whistles have been preprogrammed into the CHAT boxes in the hope that Herzing might get a chance to greet the dolphins and interact with them. It’s almost as if they’ve come to find us rather than the other way around.
Herzing has known most of her dolphins since birth, and she knows their mothers, aunts, and grandmothers as well. Many readers of this magazine know one of them too: Nereide’s mother, Nassau, appeared on the cover of National Geographic in September 1992, swimming beneath the surface of the same Bahamian waters.
These two females represent the best candidates for Herzing’s work. They haven’t yet become pregnant and are still just kids, with lots of curiosity and lots of freedom to play and explore. Sexual maturity in female Atlantic spotted dolphins arrives around age nine. Their life span can be more than 50 years.
When Herzing dives into the water and plays Meridian’s signature whistle for the first time, the dolphin turns and approaches, though without any outward sign of the surprise one might expect from a creature that’s just heard its name called by another species.
Herzing swims with her right arm stretched out in front of her, pointing at a red scarf she has pulled out of her swimsuit. She repeatedly presses the button for “scarf” on the CHAT box. It’s a rolling chirp that dips low and ends high, lasting about a second. One of the dolphins swims over, grabs the piece of fabric, and moves it back and forth from its rostrum to its pectoral fin. The scarf ends up hanging from the dolphin’s tail as she dives down to the bottom of the ocean.
I’m in the water with Herzing, trailing a few feet behind her with a graduate student who’s recording the encounter using an underwater camera. I keep waiting for one of the dolphins to take off with the scarf, but neither of them does. They seem to want to engage us, however tentatively. They pass the scarf back and forth, circle around us, disappear with it, and then offer it back to Herzing. She grabs it and tucks it back into her swimsuit and then pulls out a piece of seaweed. Nereide swoops down to grab it between her teeth and starts to swim off. Herzing takes off after her, pressing the CHAT box’s sargassum whistle again and again, as if desperately asking for it back. But the dolphins just ignore her.
“It’s not inconceivable that if the dolphins understand that we’re trying to use symbols, that they would try to show us something,” Herzing says later, back on board the Stenella. “Or imagine if they started using our word for sargassum amongst themselves.”
For now that still feels like a distant dream. The CHAT box never registers any mimicking during this hour-long encounter. “It’s all about exposure, exposure, exposure,” says Herzing. A tall order when you’re a human on a boat trying to link up with wild dolphins for a brief chat in a vast ocean.
“They’re curious. You can see them starting to put it together. I just keep waiting for them to trigger,” she says. “I keep waiting to hear a female voice in my headphones saying, ‘Scarf!’ You can almost see them calculating in their eyes, trying to work it out. If only they’d give me some acoustical feedback.”
The feedback may be there, just not in a form anyone can make sense of yet. Nereide had draped the sargassum over her tail as she floated casually through the water, finally shaking it off and then blowing a big, playful bubble.
After an hour in the water with us, the dolphins began to lose interest. As Nereide turned to leave, she made one final long, mysterious whistle, looked back at us, and then swam off into the blue darkness and disappeared.
By Joshua Foer
Photographs by Brian Skerry
National Geographic, May 2015