They live in multicultural, matriarchal societies. Their 18-pound brains have neocortexes and spindle cells for higher-order thinking, emotions, memory, language, and love. They could be the most intelligent beings on this planet. We're talking about sperm whales. And by the way, they don't just make whale sounds: They speak.
According to City University of New York biological oceanographer David Gruber, sperm whale vocalizations aren't your typical harmonious whale songs. Their phrases-series of Morse Code-like clicks called codas-sound more like a door slowly creaking open and shut on the deck of a sunken boat. But when researchers really started listening, they found that these codas carry all the hallmarks of a highly evolved language, even regional dialects, passed down generation to generation. In fact, Gruber has said he believes sperm whale speak to be "perhaps the most sophisticated form of communication that has ever existed."
In 2021, Gruber officially launched the CETI Project (Cetacean Translation Initiative), a global interdisciplinary collaboration devoted to cracking the code on interspecies interpretation and communication, starting with the resident population of sperm whales in Dominica, in the Eastern Caribbean. The list of partner institutions includes the Dominica government, and about 50 cryptographers, linguists, technologists, and biologists who are heavyweights in their own fields. There's natural language processing expert Michael Bronstein from Oxford University, Harvard University roboticists who specialize in extremely gentle technology for humane animal research, and Roger Payne-the American biologist who, 56 years ago, discovered for the first time that humpback whales sing, sparking the marine conservation movement. Tech superpowers like Google, Amazon and Microsoft are on board, too.
The end goal of the project: an open-source database and interspecies communication toolset that can be applied one day to other species, too-on Earth and. who knows where else?
CETI recently raised a significant chunk of funding under the umbrella of TED's Audacious Project-an incubator program for world-changing ideas. At the 2023 TED conference, Gruber held an intimate learning session where he presented the latest on the project's progress. According to Gruber, when Payne first began studying humpback songs in the 1960s, he didn't bring human language experts into the mix. This time around, the plan is different: "Linguists are going to be the special sauce," he told attendees. "It's almost like Arrival."
Interspecies listening and translation research is happening across the animal kingdom, from birds and bats, to turtles, to honeybees. But once humans understand other species, the next logical step is to talk back. That, according to Canadian scientist and Rhodes Scholar Karen Bakker, opens up a whole can of worms.
"Some believe that interspecies communication would help foster respect and empathy for nature," Bakker, a professor at the University of British Columbia who studies bioacoustics and non-human communication, said in a yet-to-be-published talk also at TED. "Others believe that it is profoundly disrespectful and unethical to eavesdrop and engage in this way."
One of the upsides, so far, she said, is the bioacoustic tech already put into use last year to triangulate the locations of massive, majestic North Atlantic right whales, of which fewer than 350 remain in the world, and convey those coordinates to ship captains in real time so they can slow down or stop. "Not a single right whale has died in a ship strike in this zone since this program was launched," she said.
Meanwhile, another group of scientists working with a non-profit called the Earth Species Project are using algorithms to generate hypothetical humpback whale vocalizations in a guess-and-check approach to understanding what certain humpback sounds mean.
The CETI team isn't ready to jump to this step, Gruber told The Daily Beast. He wants his team to master listening-and to feel confident that they understand to some extent-whale speak before they fire off messages into the deep ocean.
"I'm coming at this from someone who's studied animals for two decades, and, I'm coming at it as someone who has also seen coral reefs disappear, quite rapidly, in my lifetime," he said, noting that it's important to him that this innovation is being led by experts specifically from the animal behavior and conservation space-not big tech. "The fact that we want to hear what the whale says is most important to me."
To do that, they're learning from scratch, like a baby would, he said, with a focus on mothers and their calves.
"Basically, we need to train the computer to be a baby whale," he said. Of course, that will take massive amounts of data-and acquiring it won't be easy.
Thanks to longtime whale-listening researcher Shane Gero, the founder of the Dominica Sperm Whale Project and a CETI collaborator, the team has an initial few thousand sperm whale vocalizations from Dominica to work with. But for CETI's custom machine learning program to do its thing, the team will soon require hundreds of millions of audio recordings, at the least. And those all need to be annotated and mapped to behaviors, down to the millisecond.
That's why CETI's work right now is mostly unfolding not on the deck of their 40-foot research sailboat in Dominica, but in the robotics lab at Harvard, where roboticists are developing and testing CETI's core whale listening station-floating beacons strung with a multitude of dangling, waterproof microphones on long, long cords-and a series of whale recording units.
"We're building all the technology de novo," Gruber noted. "It's not like you just go get off-the-shelf equipment on how to communicate with whales."
Complicating matters further is the team's commitment to never drawing a single drop of blood from its study subjects. That makes it a big challenge to tag whales with trackers, mics, and eventually cameras that stay put on animals that regularly dive as deep as a mile underwater.
When data is collected (about four terabytes per month), it's uploaded from the Eastern Caribbean to the cloud, and CETI's algorithms start searching for linguistically significant patterns, like recursion (which requires grammar) or displacement (the discussion of things that are not immediately present, physically or in terms of time-only observed in human communication so far).
Eventually, the researchers will start playing back codas to the whales, not necessarily to talk to them-yet-but to test theories about what they're hearing. "In my mind, it's not really that I have something to say to a whale," Gruber says. "This is a way to validate that our algorithms are right. [...] The fact that we want to hear what the whale says is most important to me."
Sperm whales and humans can relate on some pretty fundamental issues: "We're mammals, and they're mammals. We share that in common," Gruber said. "Obviously, they're eating and they're sleeping. And they're caretaking. And they're babysitting. So there are some features that we definitely know are overlapping."
But he is most curious about the parts of our world and their world that don't overlap. "Their world is so different from ours, and those are the fun things that my brain goes to," he said.
Scientists are supposed to be "completely objective," he added, but it's pretty much impossible to shed the biases about how the world works that come from the sheer experience of being human-breathing oxygen, eating food, amassing and sharing knowledge, living in an environment with gravity.
"We have to keep open to what the data will show us, and what new things will come out," he said. "And that, I think, will be relevant if we were to meet an extraterrestrial. Honestly, the first question for me is: Are we ready to meet an alien? Say the whale was the alien: When we met it, we harpooned it."
Gruber doesn't feel confident this will become a pressing issue within his lifetime, but if and when humans do encounter extraterrestrial life, he believes there is a real risk they'd make the same mistakes they made when encountering new species on Earth. CETI, he hopes, could help. "Part of this is to get our training wheels and get ready for that.
"I work with sharks, and for 400 million years, they've been top predators. For so long that they've developed this kind of consciousness to be at the top of the food chain," Gruber said. "Humans, we're kind of mid-tier primates. We [haven't] been at the top that long. That's why I think we like to watch 'Shark Week:' because we still have that kind of primal part of our brain."
Ultimately, Gruber thinks the CETI project isn't simply an exercise in learning how to talk to aliens, but a longer practice that could teach us how to ensure these communications occur peacefully and ethically for all involved. "I think it's kind of like a race to get our consciousness to the point where we recognize how powerful we are, but also how fragile and vulnerable we are-you know, before we destroy everything."
Read more at The Daily Beast.