Complex Language: Scientists find an ‘alphabet’ in whale songs
In 2020, a team of marine biologists and computer scientists joined forces to analyze the click-clacking songs of sperm whales, the gray, block-shaped leviathans that swim in most of the world’s oceans
Ever since the discovery of whale songs almost 60 years ago, scientists have been trying to decipher their lyrics. Are the animals producing complex messages akin to human language? Or sharing simpler pieces of information, like dancing bees do? Or are they communicating something else we don’t yet understand?
In 2020, a team of marine biologists and computer scientists joined forces to analyze the click-clacking songs of sperm whales, the gray, block-shaped leviathans that swim in most of the world’s oceans. On Tuesday, the scientists reported that the whales use a much richer set of sounds than previously known, which they called a “sperm whale phonetic alphabet.”
People have a pho-ne-tic alphabet too, which we use to produce a practically infinite supply of words. But Shane Gero, a marine biologist at Carleton University in Ottawa and an author of the study, said it’s unclear whether sperm whales similarly turn their phonetic sounds into a language.
“The fundamental similarities that we do find are really fascinating,” Dr. Gero said. “It’s totally changed the way we have to do work going forward.” Since 2005, Dr. Gero and his colleagues have followed a clan of 400 sperm whales around Dominica, an island nation in the eastern Caribbean, eavesdropping on the whales with underwater microphones and tagging some of the animals with sensors.
Sperm whales don’t produce the eerie melodies sung by humpback whales, which became a sensation in the 1960s. Instead, they rattle off clicks that sound like a cross between Morse code and a creaking door. Sperm whales typically produce pulses of between three and 40 clicks, known as codas. They usually sing these codas while swimming together, raising the possibility that they’re communicating with one another.
Over the years, Dr. Gero and his colleagues have reviewed thousands of hours of recordings of the undersea noise. It turns out that sperm whale codas fall into distinct types.
One type, for example, called “1+1+3,” consists of two clicks separated by a pause, followed by three clicks in quick succession.
With backing from philanthropists, Dr. Gero and his colleagues started “Project CETI,” (for “Cetacean Translation Initiative”), to investigate whether artificial intelligence and other computing advances could decode whale songs. (The name is a play on SETI, the famous effort to search for extraterrestrial life; whales are also known as cetaceans.)
As part of the project, Pratyusha Sharma, a computer science graduate student at M.I.T., gave the data from Dominica a fresh look. But she was frustrated by the way biologists had visualized it.
On a computer screen, the codas appeared as a series of dots along a horizontal line, each dot representing a click. Sharma found it hard to compare codas, especially when two or more whales were singing over each other. So she instead plotted each coda’s clicks as dots on a vertical line, and then placed codas along a horizontal line based on when each began.
Using the new layout, Sharma saw something new. When a sperm whale repeated a coda, it sometimes stretched out the time between the clicks and then gradually tightened it up. Sharma and her colleagues called this phenomenon “rubato,” a musical term for speeding up a tempo and then slowing it down.
Dr. Gero was startled that Sharma could see something in sperm whale songs that he and his colleagues had missed for years. “It was a way we hadn’t looked at it,” he said.
Codas are so quick that the human ear can miss a rubato. But the researchers found the pattern in thousands of recorded codas.
The researchers believe that rubato plays an important role in whale communication. They found that after one whale used rubato, neighboring whales would rapidly match the tempo change with their own codas.
Sharma’s new visualizations also revealed that sperm whales could occasionally add an extra click to the end of the coda, a behavior they call ornamentation. The scientists found evidence that the extra clicks were not just pointless flourishes. The whales that led groups often used ornamentation, after which their followers often responded with codas of their own.
The analysis showed that the conventional catalog of sperm whale codas could not capture their full complexity. Sperm whales can produce a 1+1+3 coda, for example, that lasts four-fifths of a second, or one second, or 1.25 seconds. Other codas may last only one-third of a second or half a second.
All told, the researchers identified 156 different codas, each with distinct combinations of tempo, rhythm, rubato and ornamentation. Dr. Gero said that this variation is strikingly similar to the way humans combine movements in our lips and tongue to produce a set of phonetic sounds.
A single sound like “ba,” or “na” carries no semantic meaning on its own. But we can combine them into meaningful words like “banana.” The researchers raised the possibility that sperm whales might combine features of codas to convey meaning in a similar way. Other experts said the whale alphabet marked an exciting advance. But they said sperm whale codas might be more akin to music than language.
“Music can have a strong influence on emotions without it actually conveying information,” said Taylor Hersh, a bioacoustician at Oregon State University. Rubato might be one way for sperm whales to tighten their social bonds, she speculated, by matching their songs.
Jacob Andreas, a computer scientist at M.I.T. and an author of the study, said that the alphabet is allowing the researchers to dig deeper into whale songs. “Now we have gotten the machinery in place to start tackling the much more ambitious, long-term goal for Project CETI, which is trying to figure out what all of this actually means.”
Microphones deployed in the Caribbean are capturing ocean sounds 24 hours a day, and scientists are programming computers to learn how to pick sperm whale songs out from the background noise. Dr. Andreas and his colleagues are also training artificial intelligence programs similar to ChatGPT. After listening to the sperm whale songs, these models might learn to recognize not just rubato and ornamentation, but other features that scientists have missed.
The hope is that computers will then be able to compose whale songs of their own, which could then be played to the whales.
That effort leaves other experts skeptical. Luke Rendell, a marine biologist at the University of St Andrews, in Scotland, worries that the A.I. models assume whale songs are a kind of language, rather than something more like music.
“I’ve no doubt that you could produce a language model that could learn to produce sperm-whale-like sequences,” Dr. Rendell said. “But that’s all you get.”