Somewhere in the deep waters off Dominica, a female sperm whale fires a rapid burst of clicks toward her companions. The sequence lasts barely a second. To the human ear, it sounds like a stuttering zipper. But according to a study published in May 2026 in Proceedings of the Royal Society B, those clicks contain layered structural rules that parallel the way human speech organizes vowels and consonants.
The research team, led by Gasper Begus of UC Berkeley and the nonprofit Project CETI, analyzed nearly four thousand click sequences, called codas, recorded from 15 female and young whales between 2014 and 2018. Within those codas, they identified distinct vowel-like categories that combine and shift depending on context, following rule-governed patterns strikingly similar to the phonological systems underlying human languages.
Vowels in the deep
Sperm whale codas are not monotone. The researchers found that subtle timing differences between individual clicks sort into consistent clusters, which they labeled “a-” and “i-” categories. These function like vowel classes: recurring, recognizable patterns of sound rather than one-off variations. Critically, the categories do not appear in isolation. They combine with rhythmic templates and with differences in the number of clicks per sequence to form higher-order patterns, much the way phonemes slot into syllables in spoken English or Japanese.
This builds on a foundation laid in 2024, when a related team published a paper in Nature Communications introducing what they called a sperm whale “phonetic alphabet.” That earlier study established that codas carry information not just through rhythm but through how specific elements are arranged relative to one another. The new Royal Society paper extends that framework by drilling into the vowel-like dimension, adding evidence that the system is more internally structured than researchers had recognized.
The underlying recordings came from two sources: digital acoustic tags (DTAGs) attached directly to whales and hydrophone arrays deployed in the water column nearby. A separate methods paper in Scientific Reports describes how the Caribbean social groups were identified and tracked over multiple field seasons, and how the nearly four thousand codas were manually annotated with inter-click intervals extracted for fine-grained analysis. That level of methodological transparency gives the phonological claims a solid empirical base, even though the full raw audio corpus has not been made broadly accessible beyond what appears in published supplements.
Structure is not the same as meaning
Identifying rules is not the same as cracking a code. In human speech, phonological contrasts matter because swapping one vowel for another can change a word’s meaning: “bit” becomes “bat.” Whether the vowel-like distinctions in sperm whale codas similarly alter the information content of a click sequence, or whether they reflect something closer to accent variation without semantic weight, remains an open question the current data cannot resolve.
Scientists still lack a reliable mapping between specific coda patterns and specific behavioral outcomes. Some codas correlate with particular contexts, such as socializing at the surface or coordinating dives, but the link is statistical, not one-to-one. Without a robust decoding of what different structural combinations convey to the whales themselves, describing the system as a full language would outrun the evidence.
Mauricio Cantor, a marine biologist at Oregon State University who was not involved in the research, offered outside perspective on the findings. His comments, reported by The Guardian, suggest that the structural parallels are striking but that cross-population data would strengthen the case for true dialect-like variation among whale clans. His caution reflects a broader pattern in the field: excitement about the complexity being uncovered, tempered by awareness that the sample represents a single community in one corner of the Caribbean.
One population, one ocean
The biggest limitation is geographic scope. The dataset covers whales from social units near Dominica. Sperm whale clans in the Pacific, Indian Ocean, and elsewhere produce different coda repertoires, and no one yet knows whether those repertoires share the same vowel-like layering or follow entirely different organizational principles. Without comparative data from multiple ocean basins, the phonological framework could describe a local communication system rather than a species-wide one.
Temporal stability is another unknown. Cultural transmission in sperm whales is thought to occur across generations, with calves learning coda types from their social group. But the phonological analysis captures a single multi-year window. Whether the same “a-” and “i-” categories will persist unchanged decades from now, or whether they drift and merge as groups interact, is a question only long-term monitoring can answer.
A separate, not-yet-peer-reviewed preprint has attempted to model how coda structure might be transmitted culturally between clans, developing computational approaches for tracking micro-level style differences as they propagate. That work offers useful methodological context, but its conclusions should be treated as preliminary until independent replication and formal review are complete.
What shifts from here
For decades, cetacean communication research consisted largely of cataloging coda types and mapping them to populations. The new work represents a different kind of question: not which clicks whales make, but what internal architecture governs how those clicks are assembled. That shift matters because it opens the door to testing whether different populations use different rule systems, whether those systems evolve over time, and whether structural contrasts carry distinct meanings in the animals’ social lives.
Project CETI, the research consortium behind much of this work, has been deploying machine-learning tools alongside traditional bioacoustic methods to parse whale vocalizations at scale. As new datasets from other oceans come online and as researchers pair high-resolution acoustic analysis with detailed behavioral observation, the picture of what sperm whales are communicating will sharpen. For now, the takeaway is precise and remarkable on its own terms: the clicks of at least one well-studied whale community are organized by layered, rule-governed patterns that, until recently, scientists assumed were unique to human speech.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
