Researchers using machine learning have discovered that the complex click patterns of sperm whales, known as “codas,” exhibit structural similarities to human speech. A new study from Project CETI highlights parallels in phonology, suggesting a level of communicative complexity previously undocumented in animal communication systems.

Whale Communication Mirrors Human Phonology

The study, published in the Proceedings of the Royal Society B, analyzed thousands of codas from sperm whales in the Eastern Caribbean. Researchers identified distinct categories of clicks that behave like vowels in human speech, with consistent differences in length and patterns. According to the paper, these vocalizations represent “one of the closest parallels to human phonology of any analyzed animal communication system,” suggesting an independent evolution of complex communication. The findings build on earlier work that first identified the acoustic resemblance of these sounds to human vowels.

The Role of AI in Decoding Cetacean Language

A central goal of Project CETI is to use machine learning to analyze sperm whale communication in detail, moving beyond pattern recognition to potentially understand the meaning of their vocalizations. This effort is similar to other initiatives where researchers use an AI toolkit for scientific research, applying advanced computational power to complex biological data. Understanding complex systems, whether in nature or in cyber warfare, requires sophisticated analytical tools. If successful, this approach could determine whether sperm whales possess a true language, capable of conveying structured, meaningful messages.

About the Author