How First Contact With Whale Civilization Could Unfold
If we can learn to speak their language, what should we say?
![A diver descends toward the head of a sperm whale swimming perpendicular to the surface.](https://cdn.theatlantic.com/thumbor/SWMCofc6K41DOpDT7F7penIiG8c=/0x0:2400x3000/648x810/media/img/2024/02/23/ShawnHeinrichs_20221105_001522_DSC9212_Edit_Edit/original.jpg)
One night last winter, over drinks in downtown Los Angeles, the biologist David Gruber told me that human beings might someday talk to sperm whales. In 2020, Gruber founded Project CETI with some of the world’s leading artificial-intelligence researchers, and they have so far raised $33 million for a high-tech effort to learn the whales’ language. Gruber said that they hope to record billions of the animals’ clicking sounds with floating hydrophones, and then to decipher the sounds’ meaning using neural networks. I was immediately intrigued. For years, I had been toiling away on a book about the search for cosmic civilizations with whom we might communicate. This one was right here on Earth.
Sperm whales are the planet’s largest-brained animals, and their nested social structures are immense. About 10 whales swim together full-time as a unit. They will sometimes meet up with others in groups of hundreds. All of the whales in these larger groups belong to clans that can contain as many as 10,000 animals, or perhaps more. (The upper limit is uncertain, because industrial whaling reduced the animals’ numbers.) Sperm whales meet just a fraction of their fellow clan members during their lifetime, but with those they do meet, they use a clan-specific dialect of click sequences called codas.
I recently visited the paleontologist Nick Pyenson in his office at the end of a long corridor of fossils at the Smithsonian Museum of Natural History. As we hefted a sperm whale’s skull out of a fiberglass crate, he told me that the clans likely date back to the Ice Age and that a few could be hundreds of thousands of years old. Their codas could be orders of magnitude more ancient than Sanskrit. We don’t know how much meaning they convey, but we do know that they’ll be very difficult to decode. Project CETI’s scientists will need to observe the whales for years and achieve fundamental breakthroughs in AI. But if they’re successful, humans could be able to initiate a conversation with whales.
This would be a first-contact scenario involving two species that have lived side by side for ages. I wanted to imagine how it could unfold. I reached out to marine biologists, field scientists who specialize in whales, paleontologists, professors of animal-rights law, linguists, and philosophers. Assume that Project CETI works, I told them. Assume that we are able to communicate something of substance to the sperm whale civilization. What should we say?
More than one person told me: Nothing at all. “I’ve always been very anti–talking to them,” said Hal Whitehead, a marine biologist at Dalhousie University, in Nova Scotia. He just wants to understand their lives while interfering with them as little as possible. Cesar Rodriguez-Garavito, a law professor at NYU who is advising Project CETI, told me that whatever we say, we must avoid harming the whales, and that we shouldn’t be too confident about our ability to predict the harms that a conversation could cause.
The sperm whales may not want to talk. They, like us, can be standoffish even toward members of their own species—and we are much more distant relations. Epochs have passed since our last common ancestor roamed the Earth. In the interim, we have pursued radically different, even alien, lifeways. The animals that birthed our shared mammal lineage crawled out of the ocean 400 million years ago. About 50 million years ago, the four-legged ancestors of whales crawled back in, likely somewhere near present-day Pakistan. We don’t know what tempted them back to the sea, but whatever it was must have had a special pull for sperm whales, which have since become the great canyoneers of the mammalian class, the seekers of some of the deepest abysses. They spend most of their lives on long dives that frequently reach thousands of feet beneath the sea surface. When humans venture that far down, we’re often encased in a thick layer of titanium, lest our lungs and ear cavities collapse. Sperm whales do it naked.
How should we even approach these strange and beguiling creatures? If we drop a speaker into the water, they might assume that our clicks are coming from an unseen clan member. A robotic whale that makes clicking sounds would perhaps fool the sperm whales’ eyes, but their echolocation beams would reveal its synthetic innards. It would be most honest to communicate in person, but whoever we’d send would need to be careful not to corner a lone whale. They would want to approach as a whole unit, so that if the whales felt threatened, they could fall into their protective rosette formation: heads in, flukes out, calves in the middle. Then we’d know to back off.
We would probably want to send a woman, on the off chance that the whales could tell the difference. (Sperm whale units are matrilineal: They’re made of grandmothers and mothers and calves. After adolescence, the males depart, and are only allowed back for sex.) Imagine that she slips on a wet suit and fins, dives into the Caribbean, and swims toward the whales with a speaker in hand. Before they see her, they will hear her. The whales’ clicks generate a bubble of perception that can extend a kilometer in every direction. As she gets closer, she might feel their echolocation beams pass through her body. From these, the whales might be able to sense her heartbeat quickening.
Over the course of their lives, they may have learned to associate the sounds of this odd, amphibious creature with many things, not all of them pleasant. They may, for instance, understand that she bears some relation to ships, the great audio polluters of their environment. The grandmother whales may associate her with the harpoons of industrial whaling.
Our messenger should come prepared with a set of clicks that might put the whales at ease. Many sperm whale dialects appear to contain a social-marker coda. It’s possible that the whales deploy it in sequence with others to express various levels of kin identity: clan, grandmother, mother. Sending out these codas may not feel like talking for the whales. It may feel more like chanting or beating an instrument in rhythm with others. Imagine the incantatory power of a drum circle spread across a huge patch of ocean. Each animal’s clicks would be its way of saying “I’m still here. We’re doing this together, you and me,” said Luke Rendell, a biologist at the University of St. Andrews, in Scotland, who has observed sperm whales for decades. Other nearby units from the clan could overhear the rhythm. It could be an invitation to join up and swim together.
Our messenger could approach the whales and play the clan’s social-marker coda through her handheld speaker. We’d have to hope that it was received as a high-minded way of saying, We are kin, which we are, if only in some distant sense. We’d have to hope that the whales would click the coda right back. But, of course, they might misconstrue its meaning. They may be horrified by a human clicking their coda. Sperm whales are the largest toothed predators in the known universe; they may wield the bludgeoning power of their flukes against her.
We might be able to add emotional warmth to our opening message, to make a misunderstanding less likely. Taylor Hersh, a postdoctoral researcher in comparative bioacoustics at Germany’s Max Planck Institute for Psycholinguistics, expects Project CETI to identify hidden nuances in the clicks. She thinks there might be some kind of animal arousal communicated in the rapidity of clicking, which could be used to convey delight. She hopes that it might be possible to say something like I am excited to see you.
At this early stage of the exchange—the first in an interspecies conversation that could end up lasting millions of years—the ultimate signal of success would be pure new-age wish fulfillment: the whales approaching the woman and barrel-rolling alongside her. Whitehead told me that last year he got drones airborne over some sperm whales in the Galápagos Islands, which gave him a new picture of how much they roll together. “They are clearly enormously important to each other,” he said. “They express this acoustically, but also with touch.” Rolling is probably too much to hope for during this initial visit, but the whales may find some milder way of conveying fellow feeling, perhaps by swimming toward our messenger in a manner that emphasizes gentleness. It would be an auspicious start.
![Two sperm whales in blue water](https://cdn.theatlantic.com/thumbor/tjMexuxuL3x4zVu_tRCNvFJkQCc=/0x0:3000x2000/928x619/media/img/posts/2024/02/ShawnHeinrichs_20221104_231114_DSC8942_Edit/original.jpg)
To engage the whales in a more extensive dialogue will be a challenge of a different order. Its difficulty has long been understood: Even when Jonah was in the whale’s belly, he did not address his captor directly. He beseeched the Lord to speak on his behalf. Project CETI is beseeching AI.
The software’s first task will be to pick codas out from the noise of the ocean and all the other sounds that sperm whales emit: the trumpeting that likely charges their echolocation chambers; the buzzes and bursts they make when they socialize; the squeals and chirps that Rendell compared to “rather intense farts.” Males sometimes fill the ocean with a mega-metronome of loud clicks that hit every six seconds. “Sailors used to tell a story about Davy Jones being stuck at the bottom of the sea and how they could hear him knocking,” Rendell said. “I’d bet money that they were hearing a male sperm whale.”
David Gruber is looking for contextual clues that will help illuminate the codas’ meanings. He’s starting with the clicks that the whales emit right before they dive, perhaps to make babysitting arrangements. Once he has amassed enough recorded codas, he wants to dump them into a neural network, not unlike the large language models that power ChatGPT. AIs are beginning to translate languages with very few translation samples. Project CETI is conducting research into one that can do it without any Rosetta stone at all, but this work is still entirely theoretical and may never escape that phase.
Assuming that at some point in the future, an AI does seem to figure out a way to translate codas, its work would still be suspect. How could we know whether the translations were accurate, or even in the ballpark? These systems are black boxes. They sometimes hallucinate. “When we do machine learning with human language, people are paid basically nothing to say whether their outputs are meaningful,” Kristin Andrews, a philosopher at York University, in Toronto, who specializes in animal minds, told me. “In this case, the whales are going to have to do the supervising.” We may be surprised to learn that what we think means Hello really means All the prey are 300 miles north; swim there fast. Our misfires may disrupt sperm whale culture, Andrews said. We could inadvertently tell them, Your leader is going to kill all of you.
Even if we can learn to speak the sperm whales’ language, other problems may arise. We don’t know that their minds are able to parse complex statements. A sperm whale’s brain is up to six times larger than ours, and its cerebral cortex is more labyrinthine. But that doesn’t necessarily make its thinking as intricate. The structure of the whales’ communication might also place severe limits on the kind of dialogue we can hope to have with them.
Scientists disagree about how much meaning the codas can express. Boosters point to the codas’ intricate structure and the fact that they are sometimes emitted in sequence. Skeptics point out that in humpback whales’ songs, as well as those sung by many birds, complexity is ornamental: No matter how melodic a song’s composition, its underlying message is often just Let’s mate. We have already decoded small sets of animal signals, including honeybee waggles and monkey alarm calls, said Dan Harris, a philosopher of language at Hunter College, but those translations didn’t kick off a rich dialogue between our species.
It’s not yet clear how many codas each sperm whale clan has in its repertoire, but all guesses are under 100. If we want to have a complex conversation, we need to hope that modulations to a codas’ speed or tone can add a great many shades of meaning. So far, Harris said, the most extensive symbolic exchange we’ve had with another animal was with a bonobo named Kanzi. We did not adopt his native modes of communication. Nor did he adopt ours. Instead, we had to build a third communication system that neither humans nor bonobos had ever used, a lexigram touch screen filled with more than 300 abstract symbols.
Richard Moore, a philosopher of language at the University of Warwick, in England, told me there is reason to believe that we could have a richer dialogue with sperm whales. He said that we have more evidence for syntax among their calls than among the gestures and sounds of bonobos, and that he wouldn’t be surprised if whales are capable of communication that is “vastly more complex” than we’re anticipating, or even comparable to human language. That doesn’t mean we can expect the whales to have the linguistic creativity of the average adult human, who knows tens of thousands of words and can arrange them into an endless variety of statements. We will always be able to say things that no whale can understand. But to say anything at all to them could be as wondrous as landing on the moon.
I asked Gruber himself what he would say to the whales. He said that he has been taking requests. Most people tell him that we should start by saying “Sorry,” for the bloody rampage that was industrial whaling. He agrees. “We pulled the oil out of these animals’ heads,” he said. “We used it to make lipstick.” Perhaps now we can atone. Charlotte Dunn, the president of the Bahamas Marine Mammal Research Organization, told me that our early communications with whales should be focused on making their lives better. She wants to ask them if ships have grown too loud and whether they would like us to move any shipping lanes. She wants to know if they can feel the climate changing. Taylor Hersh said that she would ask the whales about the worst thing in their life, in case it was something we could fix.
Before we put these questions to a sperm whale unit, we’d have to think hard about whether we’d act on the answers. Kristin Andrews told me a heartbreaking story about a chimpanzee named Bruno who was taught sign language at the University of Oklahoma. Bruno was encouraged to build his whole life around the practice of asking humans for things. But after a few years, the scientists’ grant ran out and he was transferred to a different facility. When one of the lab’s scientists visited him there, he was distressed to see that Bruno seemed upset. He kept signing Key and Out. The scientist had taught the chimpanzee to communicate, but even in the face of a clear request, the scientist couldn’t help him. “If these whales start saying Go away; make the ships leave, what will we do?” Andrews said. And how will it reflect on us as a society if we ignore them?
I asked Andrews my final question, the same one that I asked all the philosophers and scientists and linguists: Assuming that we could have a complex conversation with sperm whales, what would she say to them? “I wouldn’t want to tell them about myself or about us,” she said. “I would want to know how they saw the world. What could they tell us that we couldn’t understand without knowing their language? I’d ask them: What are you interested in? What do you value? To you, what is the meaning of life?”
Nick Pyenson, the paleontologist with the sperm whale fossils in his office, told me that he’d ask the whales about their first memory. Right after calves are born, their mothers take them to the surface to breathe. He wants to know if they remember that. Richard Moore would ask the whales how they navigate the immensity of the ocean. Hal Whitehead would ask about oceanography, about changes in currents. He’d ask them how many of their preferred prey—giant squid—populate the pitch-black depths where they hunt. Diane Claridge, the executive director of the Bahamas Marine Mammal Research Organization, would want them to describe what it’s like to swim through a hurricane.
We have knowledge of our own to share with the whales. We could catch them up on all that they’ve missed on dry land. We could tell them that right after they returned to the sea, the Himalayas began to rise and their fellow mammals diversified across the continents. We could tell them what human civilization has come to know about nature: that the sun is only one star among countless others; that we have peered billions of light-years into the universe without seeing its end. This last bit may not even surprise them. As animals of the deep ocean, they may be primed to understand the cosmos as a dark and mysterious expanse.
We could tell them about Voyager 1, the most distant probe we have sent into space. We could tell them that a golden record is affixed to its side, a message for any civilization that intercepts the spacecraft. We could tell them that etched into its surface is a map of Earth’s location in the Milky Way. The sperm whales, being creatures of sound, may be delighted to hear that the record also contains 27 audio tracks. We could tell them that whoever plays it will hear heartbeats and laughs, greetings in different languages, symphonies and folk tunes, and also the clank of tools, the rolling of thunder, the crashing of waves. And we could tell them that one of the first tracks on the golden record ends with a segment of humpback whalesong, because when we set out to identify ourselves to the universe, when we sent our coda out into its depths, one of the first things we wanted its recipients to understand is that in our clan, there are whales.