Groundbreaking Study Reveals Sperm Whale Clicks Mirror Human Speech Patterns

Sperm whales produce a series of rapid clicks, known as codas, that researchers have now shown contain vowel‑like elements and phonetic rules akin to those of human speech.

Using advanced acoustic analysis and artificial‑intelligence tools, a team led by linguist Gašper Beguš at UC Berkeley found that variations in click length, pitch rise, and fall encode distinct “vowel” sounds, creating patterns comparable to languages such as Mandarin, Latin and Slovenian.

The findings, published in the Proceedings B journal, describe the whale communication system as “highly complex” and one of the closest animal parallels to human phonology, indicating a case of independent evolution of language‑like structures.

The research was conducted by Project CETI (Cetacean Translation Initiative), which has been recording sperm whales off Dominica. The project recently released video of a collaborative birth, underscoring the species’ rich social lives.

According to Project CETI founder David Gruber, the whales’ “chit‑chat” occurs when individuals press their heads together near the surface, a behavior he likens to intimate, face‑to‑face conversation rather than distant shouting.

By removing silent gaps between clicks, the team uncovered rhythmic patterns that function like human vowel modulation—altering vocal fold tension to shift an “A” into an “E.” This level of linguistic sophistication surpasses that observed in other vocal animals such as parrots and elephants.

Behavioral ecologist Mauricio Cantor (not involved in the study) noted that the discovery reveals multiple interacting layers of structure in whale signals, a complexity previously unappreciated.

Project CETI aims to identify at least 20 distinct vocal expressions—covering actions like diving, sleeping, and social bonding—within the next five years, moving toward a functional understanding of cetacean communication.

Gruber remains optimistic, comparing current progress to a two‑year‑old child speaking a few words, and hopes that future research will bring the field to a five‑year‑old level of linguistic capability.