by Sign up for CNN’s Wonder Theory science newsletter. Explore the universe with news on exciting discoveries, scientific advances and more.
Scientists have discovered a whale. They identified a previously unknown complexity in whale communication by analyzing thousands of recorded sequences of sperm whale clicks with artificial intelligence.
Variations in the speed, rhythm and length of the whales’ click sequences, known as codas, weave a rich acoustic tapestry. These variables suggest that whales can combine click patterns in a variety of ways, mixing and matching phrases to communicate a wide range of information to each other.
What sperm whales are saying with their clicks is a mystery to human ears. However, revealing the scope of whale vocal exchanges is an important step toward linking whale calls to specific messages or social behaviors, the scientists reported May 7 in the journal Nature Communications.
“This work builds on a lot of previous work aimed at understanding the calls of sperm whales. However, this is the first work that has begun to look at sperm whale calls in their wider communication context and in the context of the exchanges between whales, which made some of the results possible,” said study coauthor Dr. Daniela Rus, director of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT, in an email.
“It helps us understand what aspects of their code they can control and change, how they can encode information in their calls,” Rus said.
The researchers called their catalog of sound combinations for sperm whales a “phonetic alphabet,” comparing variations in the whales’ click sequences to the production of different phonetic sounds in human speech.
But while the team’s findings are interesting, that term gives a misleading perspective on whale vocal interactions, said Dr. Luke Rendell, a researcher at the University of St. email.
“Presenting the ‘phonetic alphabet’ — it’s nothing,” said Rendell, who was not involved in the research.
“The way the change of pace is used is completely different from the way, say, we use elements of the alphabet to construct a language representation,” he said. “There’s no evidence for that, and it’s not a very helpful interpretation because it puts everything into a narrow and oversold perspective of ‘is it like human language or not,’ when there’s a much wider range of interpretations available.”
Pattern recognition
Sperm whales produce their clicks by forcing air through an organ in their head called the spermaceti, and these sounds can be as loud as 230 decibels – louder than a rocket launch and capable of tearing human ears – another team of scientists previously reported in the journal. Scientific Reports.
For the new study, the researchers used machine learning to detect patterns in sound data collected by The Dominica Sperm Whale Project, a repository for observing sperm whales living in the Caribbean. The recordings featured the vocalizations of approximately 60 sperm whales – a subset of approximately 400 whales known as the Eastern Caribbean clan – and the vocalizations were recorded between 2005 and 2018.
Previous research has identified 150 types of codas in sperm whales around the world, but only 21 of these codas were used by Caribbean whales.
The scientists examined the timing and frequency of 8,719 part sequences — in individual whale calls, in calls and in call-and-response exchanges between whales. When exposed to artificial intelligence, never-before-seen part patterns emerged.
The authors of the study defined four elements in codas: rhythm, tempo, rubato and ornamentation. Rhythm describes the sequence of intervals between what happens. The speed is the length of the entire code. Rubato refers to variations in tempo across adjacent parts of the same rhythm and tempo. And an ornament is an “extra click” added at the end of a section in a shorter section group, Rus explained.
These so-called ornamental clicks “occur more toward the beginning and end of turns” during vocal exchanges between whales, “behaving as discourse markers,” Rus said.
“The discovery that whales can synchronize differences in tempo code was a very interesting observation,” said Rendell.
“The ‘decoration’ bothered me less,” he said. “It rarely happens, and I think we need more evidence that it’s not just production glitches,” or filler sounds, “like when we say ‘um’ or ‘err’.”
In total, 18 types of rhythm, five types of speed, three types of rubato and two types of ornamentation were found in the program. All of these part features could be mixed and matched to create a “massive storehouse” of phrases, the study authors reported. Furthermore, meaning could be altered even more depending on the placement of a part — following or overlapping another part — within an exchange or chorus involving two or more whales.
An interactive experiment
“Actually, many of us have been waiting for advanced technology to allow us to do something like this for years!” said Dr. Brenda McCowan, a professor at the University of California Davis School of Veterinary Medicine, in an email.
McCowan, who was not involved in the research, was part of another team that conducted an interactive “conversation” in 2021 with a humpback whale in waters off Alaska. For about 20 minutes, a curious whale repeatedly responded to a humming song recording transmitted from the scientists’ boat.
“This particular replay (with the humpback in 2021) was an opportunistic experiment with a curious whale engaging us behaviorally and vocally, and entirely of its own volition,” McCowan said.
Such an interactive experiment with whales, as well as observations of whale behavior, could be an important part of solving the syntax of sperm whale click sequences, the authors wrote in the study.
Their machine learning method could also be useful for studying other types of animal vocalizations, McCowan added.
“Speed, rhythm, rubato and ornamentation are likely to be found in other whale species,” McCowan said. “We already know this to be true of the daffodil’s song. But there is also evidence of this type of pattern in other aquatic, terrestrial and tree species to which this approach could be applied.”
But while this technique is helpful in identifying certain aspects of communication, it is not a Rosetta stone, Rendell cautioned.
“Machine learning is great for finding patterns in large data sets,” he said, “but it doesn’t create meaning.”
Mindy Weisberger is a science writer and media producer whose work has appeared in Live Science, Scientific American and How It Works magazine.
For more CNN news and newsletters create an account at CNN.com
