Sperm whale coda dimensions · acoustic structure only
Sperm whales produce codas — sequences of clicks lasting one to two seconds. A 2024 study found codas vary along four independent dimensions simultaneously: rhythm (interval pattern), tempo (overall speed), rubato (systematic compression or extension), and ornamentation (extra clicks). A 2025 study added a fifth: vowel quality (spectral content, analogous to formants). The dimensions combine combinatorially, generating a much larger space of distinct codas than previously counted.
This page lets you explore those dimensions acoustically. The sounds are synthetic approximations — click sequences generated in the browser by the Web Audio API, shaped to roughly match the described properties. What you hear is structure, not meaning. The structural findings are real; what any coda refers to is a separate question that remains open.
Dimension 1
Rhythm
Pattern of inter-click intervals — whether clicks are evenly spaced, grouped, or follow a specific sequence.
Dimension 2
Tempo
Overall rate — how fast the full coda sequence is delivered, independent of the rhythm pattern.
3.0×
Dimension 3
Rubato
Systematic stretch or compression — whether intervals expand (rallentando) or compress (accelerando) across the coda duration.
0
Dimension 4
Ornamentation
Extra clicks added between the main sequence — each gap has this probability of containing an ornamental click.
0%
Dimension 5
Vowel Quality
Spectral content of each click — the 2025 study found two discrete vowel-like patterns (a-coda / i-coda) distinguishable by the frequency properties of the clicks, analogous to vowel formants. Both can appear on any rhythm pattern.
Click sequence
Click play to hear the current settings
This demonstrates acoustic structure — the combinatorial dimensions the 2024 and 2025 research identified. It does not demonstrate meaning. The same coda structure could function as an identity signal, a location marker, a social cue, or something with no referential content at all. Distinguishing these would require knowing what was happening in the whale's environment when each coda was produced — a behavioral dataset that doesn't yet exist at the required resolution. See entry 538 for the full context.