How Whales Sing: The Sonic Architecture of Cetacean Communication

The longest sustained vocalization of any animal is the song of the male humpback whale. In a typical breeding-season performance a male hangs vertically in 30 meters of water, head down, tail near the surface, and produces a sequence of moans, cries, and chirps for somewhere between 10 minutes and 30 minutes. He pauses, surfaces to breathe, dives, and produces the same sequence again. He repeats this sequence for hours. He repeats it for days. The other males in the same population are producing nearly the same sequence. The agreement is not approximate: at any given moment, the song of a humpback in Hawaii will have units, themes, and overall structure that match the songs of every other singing humpback in Hawaii, with perhaps one or two units in transition.

The song is not innate. Year over year it changes, and the changes propagate through the population in a coordinated way that ethologists describe as cultural transmission. Within a season the song stays largely stable, with small revisions; between seasons it can shift dramatically. In 1996 Michael Noad and colleagues observed something extraordinary in Australian waters: the East Australian humpback population, which had its own distinctive song, was almost entirely overrun in a single year by the song of the West Australian population. Two whales had apparently arrived from the western group and brought their song. Within months the eastern males were singing the western song almost note-for-note, and the original eastern song had nearly disappeared. Researchers had documented horizontal cultural transmission of an animal song on a population-wide scale.

Roger Payne and the 1971 recording

The science of whale song is younger than the awareness of it. Whales had been hunted for centuries; their vocalizations were known only as occasional underwater rumblings noted by submarine sonar operators and Navy hydrophones. In the 1960s Roger Payne and Scott McVay, working from Navy hydrophone data, began analyzing the vocalizations of humpbacks. They observed that the sounds had structure: discrete units repeated in the same order, themes built from sequences of units, songs built from sequences of themes. Their 1971 paper in Science, "Songs of Humpback Whales," made the case that what they were hearing was a genuine song in the ethological sense: a learned, structured, culturally transmitted vocal performance.

That same year Payne released the LP Songs of the Humpback Whale on CRM Records. The album sold over a hundred million copies. National Geographic included a flexi-disc of the recording in its January 1979 issue, distributing it to over ten million subscribers. The cultural effect was immediate. Whale songs became a fixture of meditation tapes and ambient music. They were broadcast into space on the Voyager Golden Record in 1977. The International Whaling Commission's 1982 moratorium on commercial whaling, which entered effect in 1986, would have been impossible without the public sentiment that Payne's recordings helped create. The argument shifted from a technical fisheries-management debate about sustainable harvest to a moral argument about creatures complex enough to sing.

The structure of the song

Subsequent decades have refined the picture. The song is hierarchically structured. The smallest unit is a vocalization lasting 1 to 4 seconds, a chirp or moan or cry of a particular spectral shape. Units combine into phrases (sequences of 2 to 6 units), phrases repeat to form themes (sequences of identical phrases that last 10 to 60 seconds), and themes string together into songs that last 10 to 30 minutes. The order of themes within a song is largely fixed for a given population, with occasional rearrangement. The repetition of phrases within themes is the source of the trance-like quality of whale song; the same phrase may repeat 5 to 20 times before the next theme begins.

The song is sex-specific. Only males sing, and they sing during the breeding season at low latitudes (Hawaii, Mexico, the Caribbean, Tonga, Australia's east and west coasts). Outside the breeding season the songs largely cease, though males have been recorded singing fragments during the migration to and from the feeding grounds. The function is widely interpreted as sexual: a courtship display, or a male-male signaling system, or both. The acoustic energy is enormous. A singing humpback can be heard underwater at distances of tens of kilometers, and in deep ocean conditions before the rise of shipping noise, possibly hundreds.

The 52 hertz whale

The most famous individual cetacean of the late 20th century may be the so-called "52 hertz whale," recorded repeatedly by Navy SOSUS hydrophones since 1989 and described in a 2004 paper by Mary Ann Daher and William Watkins. The whale's calls have a fundamental frequency of about 52 Hz, well above the 15 to 25 Hz typical of fin whales and the 20 Hz of blue whales. No other whale on the SOSUS records calls at that frequency. Popular accounts described the animal as "the loneliest whale in the world," singing at a pitch no other whale could understand.

The reality is probably more complicated. Some researchers have suggested the 52 Hz signal could represent a hybrid between blue and fin whales. Others have noted that the calls have shifted slightly downward in frequency over the decades (from about 52 Hz in 1989 to around 47 Hz by the 2010s), which is consistent with normal growth and aging of an individual baleen whale. The animal continues to be recorded, and a 2010 Long Beach signal was actually detected by multiple sensors simultaneously, suggesting either two animals calling at similar frequencies or a single animal whose calls travel widely. The "loneliest whale" framing is sentimental; the underlying biology, that an individual whale calls at an unusual frequency for reasons we do not understand, remains genuinely mysterious.

How whales make sound

Toothed whales (sperm whales, dolphins, beaked whales) produce sound differently from baleen whales (humpbacks, blues, fins). Toothed whales have a structure called the phonic lip in the nasal passage, which is forced into vibration by air moving through it. The sound is then focused by a fatty organ called the melon, in the forehead, into a directional beam. Dolphins use this system for echolocation; sperm whales use a much larger version to produce the loudest biological sounds on Earth, click sequences exceeding 230 decibels at the source.

Baleen whales, including humpbacks, lack the phonic lip. They produce sound by moving air through their respiratory system, but exactly how the song is generated remains incompletely understood. A 2006 anatomical study by Reidenberg and Laitman identified a "U-fold" structure in the larynx of baleen whales that they proposed as the vocal source. The whale would not need to exhale through this structure; air could be circulated within closed nasal sacs, allowing sustained vocalization without needing to surface for breath. This is consistent with what is observed: humpbacks can sing for 20 minutes between breaths, longer than they could plausibly sustain by exhaling.

The shipping noise problem

The acoustic environment in which whales evolved no longer exists. Industrial shipping has raised the ambient noise floor in the world's oceans by roughly 3 decibels per decade since the 1960s. The frequencies that overlap most with whale calls (20 Hz to a few hundred Hz) are dominated by ship engines and propellers. A study by Christopher Clark and colleagues estimated that the communication range of fin whales has shrunk from approximately 1500 kilometers in the pre-industrial ocean to about 50 kilometers in the busiest modern shipping lanes. The whales are still calling, but they may no longer be heard by the whales they are calling to.

Some species have responded by shifting their calls to higher frequencies, above the dominant ship-noise band, or by calling more loudly. North Atlantic right whales, one of the most endangered species, have been documented raising the amplitude of their calls in response to increased ambient noise, the marine equivalent of speaking up at a loud party. Whether the long-term population effect is reproductive failure (because mates cannot find each other), behavioral disruption (because the energy budget shifts toward calling more loudly), or some combination, is an active area of research.

The story of whale song is in some ways a story about the limits of our attention. For centuries we hunted these animals without knowing they sang. For half a century we have known they sing and have done relatively little to preserve the acoustic environments in which their songs make sense. The songs themselves continue to evolve, the cultural transmission continues to surprise us, and the deeper questions about what the songs mean to the singers remain mostly unanswered. The most reliable thing we know about whale song is that we know less about it than we used to think.