Whale Song
A Cold War hydrophone caught it by accident, a record album carried it into ten million households, and nobody yet knows what it means.

In 1967 a Navy engineer named Frank Watlington took Roger and Katharine Payne out on a boat off Bermuda, threaded a reel of tape onto a recorder, and played them something he had been sitting on for a decade. Watlington ran a hydrophone array for the American underwater surveillance system, a chain of microphones anchored on the sea floor to listen for Soviet submarines, and the same unplaceable sounds kept intruding on his tapes. He had a guess. “I think these are humpback whales,” he told Payne. Payne said afterwards that the sound “completely blew my mind.” He had never, he said, heard anything in nature “remotely as extraordinary a performance.”
He took copies back to New York and left them playing for weeks, over and over, until the thing that turns a recording into a discovery finally surfaced. “My God,” he remembered thinking, “this thing is repeating itself.”
Payne and the Princeton researcher Scott McVay published the finding in Science on 13 August 1971. Humpbacks, they wrote, emit “a series of beautiful and varied sounds for a period of 7 to 30 minutes and then repeat the same series with considerable precision.” A fixed sequence, reproduced deliberately, is the definition of a song, and no other animal was known to sustain one at that length. The paper ended on an admission the field has never fully retired: the authors could not say which sex was singing, or why.
The architecture turned out to be nested, and holding it in the head is what separates whale song from birdsong in conversation. The smallest element is a unit, one continuous sound lasting about a second. A few units in a set order make a phrase, fifteen to twenty seconds long. A phrase repeated over and over becomes a theme, running a couple of minutes. Four to six themes in a fixed sequence make one song, twelve to fifteen minutes end to end, and a male will cycle that song for hours without an obvious break. Only males have ever been observed singing, and they do it on the winter breeding grounds, not on the summer feeding grounds where most people see whales.
Wait for the repeat. A phrase runs fifteen to twenty seconds, and the moment it comes round again in the same order is the moment Payne heard in his apartment. Underneath the high moans there is a low pulsed rasp that carries furthest through water; a singer in the Hawaiian channels between December and April can be the loudest thing in the ocean for miles in every direction.
Katharine Payne, working the same Bermuda tapes, found the part that still startles people. The song is not fixed. Every male in a population sings the same version at the same time, and that shared version drifts through the season and goes on moving between seasons, so that within a few years no whale is singing what he sang before. Then, in 1996, a handful of singers from the Indian Ocean population off western Australia turned up on the Pacific side of the continent. Within two breeding seasons the entire east coast population had abandoned its own song and taken up theirs. Michael Noad and his colleagues reported it in Nature on 30 November 2000 and called it a cultural revolution, meaning the word precisely: the replacement was rapid, total, and unmatched in any other animal tradition on record.
Payne decided almost at once that the tapes had to leave the laboratory. “I just decided the world has to hear this,” he said. Songs of the Humpback Whale appeared in 1970 with Watlington credited beside the Paynes, and it has sold more than 125,000 copies, which Guinness still certifies as the biggest-selling nature album ever made. That figure badly understates the reach. In January 1979 National Geographic bound a flexible plastic disc of the songs into the magazine and mailed it to 10.5 million subscribers, which the Society describes as the largest single pressing ever made of that kind of record. Two years earlier a four-minute cut of the whales had gone aboard both Voyager probes on the Golden Record, sharing track three with the greetings of the United Nations. Both are now in interstellar space.
The politics moved with the sound. Whaling fleets were still killing tens of thousands of large whales a year when the record came out, and in 1970 Payne carried the recordings into hearings at the Department of the Interior and played them to the officials instead of arguing. Congress passed the Marine Mammal Protection Act in 1972 and the Endangered Species Act the following year. The International Whaling Commission voted a commercial moratorium in 1982. Humpback populations that commercial whaling had cut by more than ninety-five percent came back; NOAA has since concluded that nine of the fourteen distinct humpback populations no longer warrant listing at all.
What the song is for remains open more than half a century after Payne and McVay said so, and the best recent work has gone at the question sideways. In February 2025 the psychologist Inbal Arnon, the whale-song specialist Ellen Garland and the linguist Simon Kirby took eight years of recordings from New Caledonia and cut them into pieces the way an infant cuts speech into words, listening for the joins where one sound stops predicting the next. The frequencies obeyed Zipf's law: the commonest element turns up about twice as often as the second and three times as often as the third, the distribution every human language yet measured produces. Asked whether they had expected that, Garland answered, “Hell, no.” Such a pattern shows up wherever a system is learned and handed on rather than issued whole, and humpbacks have been passing this one down for longer than anyone can measure. That is what the statistics report, not the meaning of any single note.