Science & Tech
Out of our depth: speaking with whales
An ambitious project seeks to use artificial intelligence to speak with sperm whales. But is this even possible? What would we say — and would they want to speak with us?
- 3078 words
- 13 minutes
Nearly 20 years after launching a long-term research project on sperm whales off the Caribbean nation of Dominica, Shane Gero has known some of the whales he studies longer than he’s known his own kids. Gero, a scientist-in-residence at Carleton University in Ottawa, has dedicated his professional life to studying sperm whales — their language, family ties, culture and conservation. The culmination of this work will be the world’s first sperm whale marine reserve, which will come into effect on Jan. 1, 2025, off the coast of Dominica. Now, Gero is working on Project CETI, an international initiative to harness artificial intelligence to translate sperm whale language. As the biology lead for Project CETI, Gero hopes to learn more about these mysterious cetaceans, whose unfathomably deep dives and complex social structures hint at hidden profundities not yet decoded.
We started the Dominica Sperm Whale Project in 2005 and earned street cred early on by working closely with the whale-watch operators, bringing people to do interpretation work on the boats and creating the concept of whale-watching as an educational tour. By the time “swim with the whales” started, which changed things dramatically, we’d already begun talking to various stakeholders about voluntary guidelines to protect the whales. But it takes a huge burden of legal labour to change things, because [a 788-square-kilometre marine reserve] affects shipping, fishing legislation and more. Post-COVID, demand for swimming with sperm whales went through the roof. That’s when the government, the operators and the fishermen all came together on the idea of a reserve. It was a win-win-win: the fishermen stop losing their fishing gear that’s getting run over by ships, the operators get [the industry] regulated and controlled, and the money can be re-nationalized and spent on programming. It took the will of a lot of different groups to come together to make the sperm whale reserve happen.
The cultural part of sperm whale life has been so much more impactful for people than learning about any of the other science that we do. People really resonate with the idea of animal culture. Behaviour is what you do, but culture is how you’ve learned to do it. It’s important to us as humans, and it might also be important to another big-brained mammal that lives in a complex society. This concept opens us up to a constructive discussion about cultural diversity as a part of biodiversity, and that maybe will give us a better way of managing and conserving wildlife.
When I think of translating what whales could be saying, I think of it as one-way translation, like predictive models where you start writing a sentence, and it knows what’s going to come next. I am super curious to know what they are saying; I’ve spent enough time watching them make sounds and then do things. But I don’t have the ego to expect that they would want to tell me anything. If we’re able to predict the coda sequences [the basic units of sperm whale communication] and we can start predicting behaviour and then see that happen, that to me will be huge. It speaks not necessarily to the structure of their linguistic capacity or all these amazing things that humans have, but it speaks to an internal life in a being that’s so different from us, which is really hard to access.
As a caveat, when we started giving names to the whales, we never thought they would be in Canadian Geographic! You meet dozens and dozens of animals, and the names were really shorthand for us to remember them. We recognize them by the trailing edge of their tail. “Fingers” was the one that had two marks that looked like fingers giving the peace sign. And there’s one that I love dearly — she’s a very large older female that flukes very beautifully named “Fruit Salad” because she was the one we took a picture of while eating fruit salad. And while they have formal names now — every family has a letter and every baby born in that family gets a name with that letter — a lot of those silly nicknames just stuck. But our science has shown that individuals matter in their social networks, and we know them as individuals — some of them longer than I’ve known my kids.
There’s a family that we know really well, where one of the males named “Alan” was at an age where he should be leaving the family. But last summer, we witnessed a sperm whale being born in that family — it was incredible — and Alan was present for it, because it was clearly such a big family event. I’ve known Alan since he was a first year calf, and it made me go back and look at photos, as if it was a memory of my kids that popped up on my iPhone or something — “Aw, so great, look how small he was!” So, yeah, we know these whales.
It’s hard to find an explanation for how complex and rich sperm whale behaviour is without accepting that they have some kind of internal level of cognition. And the behaviours that we see, like group decision-making, movement and babysitting, may be mediated through sound, or may not. There’s this amazing problem with echolocation called “perception entanglement.” If I can see through your eyes and you can see through my eyes, I don’t have to ask you questions about what your room looks like. You don’t have to say, “Hey, Shane, where are you?” Because you can see out of my eyes all the time. And so, they may not have words for those kinds of questions, but that doesn’t mean that it’s lesser than the communication system we have. This work is ultimately about listening and building tools to understand what is important to something so different from us, in order to ask: how can we help?
One of the big takeaways about being a sperm whale is being a good neighbour. And humans have not been good neighbours. There are animals alive today that were alive back when we were still hunting and killing their species. And we’re really loud neighbors; we do not have noise violations after 11 p.m. at the bottom of the ocean like we have in our own cities. And there’s all of our plastics and our toxicology. So, I think [listening to them and better understanding their needs] is the first step toward being a better neighbour.
This story is from the November/December 2024 Issue
Science & Tech
An ambitious project seeks to use artificial intelligence to speak with sperm whales. But is this even possible? What would we say — and would they want to speak with us?
Wildlife
Plus: orca don’t love metal music, orangutans get new home at Toronto Zoo, Dominica protects ‘carbon heroes’ of the sea, and crickets boost acoustic efficiency in surprising ways
Wildlife
Plus: Canadian scientist witnesses sperm whale birth, wildlife get that shrinking feeling, migratory birds fly ever higher, and teeth tell time (sort of)
Science & Tech
Can a computer learn to identify species?
Science & Tech
An ambitious project seeks to use artificial intelligence to speak with sperm whales. But is this even possible? What would we say — and would they want to speak with us?
Wildlife
Plus: orca don’t love metal music, orangutans get new home at Toronto Zoo, Dominica protects ‘carbon heroes’ of the sea, and crickets boost acoustic efficiency in surprising ways