A Forgotten 1949 Recording Turns Out To Be The Oldest-Known Recording of a Whale Song - Discover Magazine

Marina: Welcome to Deep Voices — I'm so glad you're here. I'm Marina, a marine biologist and science communicator, and this is the show where we dive into the world of whales and dolphins — their songs, their cultures, their family bonds, and the extraordinary science working to understand what they're actually saying. Marina: A huge thank you to today's sponsors — Ada, who helps organizations integrate A I into their real-world workflows; Ago Consulting, pronounced "Ahh-Go," your silicon development partners from A I all the way to SoC design; and Zen Semiconductor, an umbrella company with a remarkable portfolio spanning the silicon industry. We couldn't make this show without them. Marina: And today — oh, today we have some stories that genuinely stopped me in my tracks. We're going back to 1949, to a recording that sat forgotten for decades and turned out to be the oldest known recording of a whale singing. We're exploring how A I is now being used to try to actually talk with whales — not just listen, but talk back. And we're sitting with a question that is equal parts thrilling and sobering: if we could have a conversation with a whale, what would that even mean — for them, and for us? Marina: Settle in. Let's go deep. Marina: So let's start with voices — and with a story that has this wonderful quality of being both very old and brand new at the same time. Marina: In 1949, a marine biologist named William Schevill and his colleague Barbara Lawrence were working off the coast of Maine. They lowered a hydrophone — an underwater microphone — into the waters near the St. Lawrence River estuary, and they recorded something. Haunting, low, rhythmic sounds coming from belugas. Now, that recording was remarkable enough. But what I want to tell you about is something that happened around the same time, something that only recently came back into the light. Marina: Researchers digging through archives discovered what is now believed to be the oldest known recording of a humpback whale song — a recording made in 1949. For decades, this audio sat essentially forgotten, unrecognized for what it was. And when scientists finally listened to it with fresh ears and modern understanding, they realized they were hearing something extraordinary: a humpback whale, singing, more than seventy years ago. Marina: Now — why does that matter? Because for a long time, the popular story of whale song began in the 1960s, with biologist Roger Payne, who made those iconic recordings of humpbacks that eventually became the best-selling nature album of all time and helped ignite the Save the Whales movement. And Payne's work was world-changing — absolutely. But this 1949 recording pushes the known history of documented whale song back further, into an era before we even had the cultural framework to understand what we were hearing. Marina: Think about that. Someone pointed a microphone into the ocean, and a whale sang into it, and nobody fully understood what had just happened. The song was there. The voice was there. We just didn't have the ears for it yet. Marina: And humpback song is worth lingering on, because it is one of the most astonishing things in the natural world. These are not random sounds. Humpback whales — and it's the males who sing the long, complex songs — produce sequences of sounds that are structured. There are themes, and phrases, and those phrases are nested within themes, and the whole thing can go on for hours. Researchers have found that humpbacks within the same ocean basin sing the same song, and that the song changes over time — and here's the part that blows my mind — the changes sweep across populations in a consistent direction, almost like a cultural wave. New phrases get introduced, adopted, spread. Old phrases fade. It is cultural transmission, happening in the ocean, in real time. Marina: So when we listen to that 1949 recording, we're not just hearing a whale. We're hearing one voice in an ongoing, evolving, living conversation that has been happening in the oceans for millions of years. We're the newcomers. We just started listening. Marina: Which brings us, naturally, to the question of what happens when we try to do more than listen. Marina: This is where things get both exciting and genuinely complicated. In the last few years, a remarkable project has been gathering momentum — and it's called Project CETI. CETI stands for Cetacean Translation Initiative, and the goal is breathtaking in its ambition: to use machine learning and A I to decode the communication of sperm whales. Marina: Why sperm whales? Partly because of their clicks — called codas — which are patterned sequences of clicks that sperm whales use to communicate with each other, particularly within their tight-knit family groups. These codas are discrete, structured, and varied enough that researchers have long suspected they carry meaningful information. The question is: what information? Marina: And this is where a study from UC Berkeley, connected to Project CETI, recently made waves — pun absolutely intended. Researchers analyzed a huge dataset of sperm whale codas and found something remarkable: the communication system of sperm whales shows features that are similar, in structural terms, to human language. We're talking about things like combinatorial structure — the idea that smaller elements get combined in different ways to create a much larger range of distinct signals. In human language, we combine sounds into words, words into sentences. Sperm whales appear to be doing something analogous with their clicks — varying rhythm, ornamentation, tempo, rubato — to generate a combinatorial system with far more expressive range than anyone had previously recognized. Marina: The researchers estimated that the sperm whale "alphabet" of codas could potentially encode thousands of distinct combinations. Thousands. That's not a simple call-and-response system. That's something that looks, structurally, like it has the capacity for complex meaning. Marina: Now — and I want to be careful here, because good science demands this — the researchers are not claiming they've decoded what sperm whales are saying. They are saying the architecture of the system is more complex and more language-like than we thought. The content, the meaning, is still the great unknown. That's where A I comes in. Marina: Project CETI is deploying underwater recording equipment, non-invasive biologging tags, and machine learning models to collect and analyze sperm whale codas at a scale never attempted before — we're talking about trying to gather billions of clicks. The A I systems being trained on this data are designed to look for patterns, contexts, and structures that human analysts would never have the bandwidth to find on their own. It's like — imagine trying to learn a new language by listening to billions of hours of conversation with no dictionary. A I can at least start to map the territory. Marina: And there are other efforts too. Scientists have been experimenting with A I systems that don't just analyze cetacean sounds, but attempt to generate responses in kind — to essentially get into a conversation. There's been genuinely documented research with humpback whales where scientists, using an underwater speaker to play back humpback calls, received responses — the whales engaged, approached, reciprocated. That is not translation, but it is contact. It is the beginning of something. Marina: And this is where some experts are urging us to slow down and think very carefully. Because — what does it mean to talk to a whale? If we develop the ability to transmit signals that whales respond to as meaningful, we have enormous power. And with that power comes enormous responsibility. Could such signals be used to manipulate whale behavior? Could we interfere with their social bonds, their migration, their feeding, their culture — while telling ourselves we're just saying hello? These are not hypothetical concerns. They are questions that serious researchers and ethicists are actively wrestling with right now. Marina: Some scientists argue that even well-intentioned communication could be disruptive if we don't understand the consequences. Others feel the potential — for conservation, for expanding our understanding of mind and intelligence in the universe — is worth pursuing with great care and humility. There's no clean consensus. And I think that's actually the right place to be. The fact that we're asking these questions before we barrel ahead says something hopeful about where science is going. Marina: Because here's what I keep coming back to: sperm whales live in matrilineal family groups. Grandmothers lead. Calves are raised by the whole family. When a member of the group is in distress, others gather around. These are beings with culture, with memory, with relationships that span decades. They are not objects to be studied from a safe clinical distance. They are subjects with their own lives, their own languages, their own stories. And if we are ever going to bridge the gap between our world and theirs, we have to approach that bridge with profound respect. Marina: Now I want to tell you a story. A real one. One that I find myself returning to when I need to remember why all of this matters. Marina: It happened in 2011, off the coast of California near San Francisco. A humpback whale — a female — was found tangled in a massive web of crab trap lines. She was wrapped around her torso, around her tail, around her flippers. The lines were pulling her down. She was exhausted. By the time a diver named James Moskito reached her, she was barely keeping her head above water. Marina: Moskito and a small team spent hours in the water, carefully cutting line after line. They had to work slowly and carefully — a thrashing whale could easily have killed them. But the whale, remarkably, stayed still. She seemed to understand, or at least to tolerate, what they were doing. As they worked, she watched them. At one point, Moskito described looking directly into her eye — this enormous, ancient eye — and feeling the weight of that gaze in a way that has stayed with him ever since. Marina: When the last line was cut, the whale didn't bolt. She swam in a wide circle around the divers. Then she came back. She swam to each diver, one by one, and appeared to nuzzle them — gently, deliberately. Moskito said she pushed him around a little, in what felt to him like an expression of joy. And then she swam away, free. Marina: Now — I am a scientist, and I want to be honest about the limits of what we can say. We cannot know with certainty what the whale was experiencing, or what her behavior meant in the way she would understand it. But we can say this: the behavior was deliberate, it was directed at the individuals who had helped her, and it was observed by multiple witnesses. And whether you call it gratitude or something we don't yet have a word for — it was real. It happened. A whale and a group of humans shared something in the water that day that neither of them will forget. Marina: And that story connects, I think, to everything we've been talking about. The ancient recording that sat waiting for ears that could hear it. The A I systems learning to map a language we're only beginning to recognize. The ethical questions about what we owe to beings with their own voices, their own cultures, their own inner lives. All of it circles back to this: relationship. Contact. The astonishing fact that we are not alone in having minds that reach out toward others. Marina: The oceans are full of voices. They have been full of voices for far longer than we've been here. And the most important thing we can do right now — perhaps the most important thing we've ever done — is learn to listen, with humility, with wonder, and with care. Marina: Thank you so much for spending this time with me today on Deep Voices. These stories — the forgotten recording, the A I researchers, the sperm whale codas, the ethical debates, the whale who circled back to say thank you — they remind me every single time why this work matters. The sea is speaking. We are just beginning to hear it. Marina: Deep Voices is brought to you by Ada, helping organizations bring A I into their real-world work; Ago Consulting — "Ahh-Go" — your partners in silicon development from A I to SoC; and Zen Semiconductor, a portfolio company spanning the silicon industry. Their support makes this show possible, and I'm genuinely grateful. Marina: If today's episode moved you — share it with someone who loves the ocean, or someone who needs to. Subscribe wherever you listen, so you never miss a dive. And check the link in show notes for sources, further reading, and everything we talked about today. Until next time — stay curious, stay kind to the sea, and keep listening for the voices in the deep.