The allure of whales has fueled human consciousness for millennia, making these ocean giants mysterious inhabitants of the deep sea. From the biblical Leviathan to Herman Melville’s mighty Moby Dick, whales play a key role in mythology and folklore. And although cetology, the study of whales, has expanded our knowledge of these marine mammals, especially in the last century, studying whales remains a huge challenge.
Now, thanks to machine learning, we’re a little closer to understanding these gentle giants. Scientists from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Cetacean Translation Initiative (CETI) project recently used algorithms to decode the “sperm whale phonetic alphabet,” revealing sophisticated sperm whale communication structures similar to human phonetics and communication systems in other countries. species of animals.
IN new open access study published in study shows that sperm whale codes, the brief bursts of clicks they employ to communicate, vary significantly in structure depending on the conversational context, revealing a communication system much more intricate than previously thought.
Nine thousand codes collected from Eastern Caribbean sperm whale families observed by the Dominic Sperm Whale Project have proven to be an instrumental starting point in uncovering the creatures’ intricate communication system. In addition to the data gold mine, the team used a range of pattern recognition and classification algorithms, as well as on-body recording equipment. They found that sperm whale communication was in fact neither random nor simplistic, but rather organized in a intricate, combinatorial way.
Scientists have identified a sort of “sperm whale phonetic alphabet” in which various elements that researchers call “rhythm”, “tempo”, “rubato” and “ornamentation” interact to create a wide range of recognizable codes. For example, the whales systematically modulated certain aspects of their codes based on the conversational context, for example by smoothly changing the duration of conversations – rubato – or adding additional decorative clicks. But even more remarkably, they found that the basic building blocks of these codes could be combined in a combinatorial way, enabling whales to construct a wide repertoire of distinct vocalizations.
The experiments were conducted using acoustic biological tags (specifically, so-called “D-tags”) deployed on clan whales from the eastern Caribbean. These tags captured the intricate details of the whales’ vocal patterns. By developing fresh data visualization and analysis techniques, CSAIL researchers discovered that individual sperm whales can emit different patterns of codes during long exchanges, rather than just repetitions of the same code. They say these patterns are nuanced and include diminutive differences that other whales also produce and recognize.
“We are venturing into the unknown to decode the secrets of sperm whale communication without any prior data,” says Daniela Rus, director of CSAIL and professor of electrical engineering and computer science (EECS) at MIT. “Using machine learning is important for identifying characteristics of their communications and predicting what they will say next. Our findings point to the presence of structured information content and challenge the belief among many linguists that complex communication is unique to humans. This is a step towards showing that other species have a previously unidentified level of communicative complexity that is deeply linked to behavior. Our next steps are to decipher the meaning of these messages and explore the social-level correlation between what is said and group actions.”
Whaling in the area
Sperm whales have the largest brains of any known animal. This is accompanied by very intricate social behaviors between families and cultural groups, requiring powerful communication for coordination, especially in high-pressure environments such as deep-sea hunting.
The whales owe much to Roger Payne, former CETI Project advisor, whale biologist, conservationist, and MacArthur Fellow, who played a major role in illuminating their musical careers. In a celebrated article from 1971, “Songs of humpback whales” Payne documented how whales can sing. His work later became a catalyst for the Save the Whales movement, a successful and timely conservation initiative.
“Roger’s research highlights the impact science can have on society. His discovery that whales sing led to the introduction of the Marine Mammal Protection Act and helped save several species of whales from extinction. This interdisciplinary study now brings us one step closer to understanding what sperm whales say,” says David Gruber, principal investigator and founder of the CETI project and distinguished professor of biology at the City University of New York.
Currently, upcoming CETI research aims to discern whether elements such as rhythm, tempo, ornamentation, and rubato convey specific communicative intentions, potentially providing insight into “pattern duality” – a linguistic phenomenon in which simple elements combine to convey complex meanings , which were previously thought to be unique to human language.
Strangers among us
“One of the intriguing aspects of our research is that it coincides with the hypothetical scenario of contact with alien species. It’s about understanding a species that has a completely different environment and communication protocols, where their interactions clearly deviate from human norms,” says Pratyusha Sharma, an MIT doctoral student in EECS, a division of CSAIL, and lead author of the study. “We study how to interpret the basic units of meaning in their communication. This isn’t just about teaching animals a subset of human language, but about decoding a naturally evolved communication system within their unique biological and environmental constraints. “In essence, our work could lay the groundwork for deciphering how an ‘alien civilization’ might communicate, providing insight into creating algorithms or systems to understand completely unknown forms of communication.”
“Many animal species have a repertoire of several distinct signals, but we are only beginning to discover the extent to which they combine these signals to create new messages,” says Robert Seyfarth, professor emeritus of psychology at the University of Pennsylvania, who was not involved in the study. “Researchers are particularly interested in whether signal combinations vary depending on the social or ecological context in which they are broadcast, and to what extent signal combinations are subject to observable “rules” recognized by listeners. The problem is particularly arduous for marine mammals because scientists are typically unable to see their objects or determine the detailed context of communication. Nevertheless, this paper presents tantalizing fresh details of call combinations and the principles underlying them in sperm whales.”
Sharma, Rus, and Gruber are joined by two others from MIT, both principal investigators of CSAIL and professors in EECS: Jacob Andreas and Antonio Torralba. They join Shane Gero, biology manager at CETI, founder of the Dominic Spem Whale Project and scientist-in-residence at Carleton University. This article was funded by the CETI Project through Dalio Philanthropies and Ocean EV Wade at MIT.