← all systems · see it in The System
Humpback whale song
Humpback whale song is long, complex, repetitive and structured in a nested hierarchy
Properties published scienceour proposal
Sourced fields, and our own readings.
| aliases | Humpback whale song culture |
| born at | unknown; Herman 2017 speculates song evolved as an aggregating call in the late Miocene-early Pliocene |
| born precision | speculation, no date |
| boundary | Inside: the patterned series of sounds a singer repeats (units, phrases, themes, songs, song types), the population-wide version all singers match, and its changes over time and between populations. Outside: the whales' other sounds that are not song (stand-alone social sounds, feeding calls), the whales' bodies and lives (humpback-whale walk), and the ocean that carries the sound. |
| nested level | Runs on mature male humpback whales (their song outlet and hearing channel, floors 6-7 of the whale walk); one of several cultural traits of a humpback population (song, migration, foraging), and so one case of culture; built from units arranged in a nested hierarchy; carried by sound in sea water. |
| scale | social |
Measures (18) published science
| song duration (one rendition) | 7 to 30 minutes |
| theme duration (a phrase repeated) | 2 to 4 minutes |
| unit frequency range | 20 Hz to upward of 24 kHz |
| syntax periodicities | 6-8 and 180-400 units |
| information rate of the unit sequence (upper bound) | less than 1 bit per second |
| time for gradual evolution to yield a different song | a few years |
| time for a revolution to replace a population's song | about 2 years (east Australia); as little as a few months within one season |
| revolution transit time, east Australia to French Polynesia | approximately two years |
| chance of learning from a neighbouring population on the feeding grounds (model) | 0.001 (CrI 8.02e-5 to 2.90e-2) probability |
| theme deletion probability per learning epoch (theme heard once; model) | 0.006 probability |
| share of model revolutions running east to west | 0.05 with empirical population sizes; 0.54 with equal sizes proportion |
| echo delay from a whale 5 km away (model example) | about 5 s |
| song unit source level | 150-189 (mean 173, s.d. 7.3) dB re 1 µPa at 1 m |
| detection range of song (Stellwagen Bank) | about 30 km |
| distance at which a sonar experiment cut song | approximately 200 km |
| share of days with humpback song (California Current, six years) | 34% rising to 76% share of days |
| New Caledonia song corpus | 46 singers, 214 songs, five song types, 47 themes, 1998-2006 count |
| span of the South Pacific ripple record | 11 years |
Open contentions (5) open contention
Where sources or readings disagree. Both stay on the map.
- C-736 When a system first throws away its excess energy and later learns to store it for reuse, is that one process that grew or two separate ones?
booked: The map keeps both readings side by side and rules neither. A few examples fit the 'one process that grew' reading, but it hasn't been tested.
the other reading: the toll and its dark twin as two things - C-1394 Is building culture up over generations unique to humans?
booked: The map records it where studies show it in other animals (pigeons, bumblebees) and doesn't rule human culture unique.
the other reading: cumulative culture (the parsimonious reading, Garland 2022); not cumulative: revolutions lose song material and complexity - C-1471 What is humpback song for?
booked: The map books it as a male display in the breeding season and leaves its exact purpose open.
the other reading: female attraction to individual singers (female visits 'virtually absent', Herman 2017); male-male interactions: dominance, affiliation or prospecting (Herman 2017); a lek: communal display drawing females to the aggregation (Herman 2017); long-range sonar (Mercado 2018; no echoes recorded before action); a vector recruiting colonists to new winter grounds (Herman 2017, suggestion) - C-1715 Why do whale song revolutions sweep the Southern Ocean?
booked: The map lists differences in learning, geography and contact, and a drive for novelty, ruling none.
the other reading: differences in learning biases between hemispheres (Zandberg 2021 names it); differences in geography and contact between populations (Zandberg 2021; Garland 2025); a sexually selected drive for novelty (Garland 2022, hypothesis) - C-1716 Does human noise change humpback song?
booked: The map keeps the early finding of no effect beside later studies where sonar and air guns changed singing.
the other reading: loud sounds such as dynamite blasts 'do not seem to affect' the songs (Payne & McVay 1971)
Partners and parts published science
Built from / builds (5)
Sits inside / holds (3)
Trades with (valence) (6)
- Human (Homo sapiens) via sys:acoustic-wave · Humpback whale song gain
- Music via sys:acoustic-wave · Humpback whale song even
- Humpback whale (Megaptera novaeangliae) via sys:acoustic-wave · Humpback whale song gain
- Naval sonar and ship noise via sys:acoustic-wave · Humpback whale song loss
- Natural ambient noise (wind) via sys:acoustic-wave · Humpback whale song loss
- Forage (krill and forage fish) in feeding habitat via sys:sensory-environment · Humpback whale song gain
expressed by (10)
- Generation/Synthesis
- Recombination/Composition
- Templating / Standardization
- Interoperability/Transduction (Bridging)
- Amplification/Scale-Up (Manufacture & Distribution)
- Distributed Autonomy/Decentralized Execution (Subsidiarity)
- Uncertainty Reduction/Learning
- Social Learning / Public Information Use
- Reactive Response
- Quality Control & Error Correction
floor filled by (11)
- Acoustic wave (sound: pressure waves in matter)
- Human (Homo sapiens)
- Ocean
- Culture
- Humpback whale (Megaptera novaeangliae)
- Humpback whale brain
- Humpback song type
- Song units
- Phrases
- Themes
- Humpback whale larynx (U-fold, laryngeal sac)
represents (1)
Kinds (8)
- Animal culture
- Cetacean culture
- Song revolution (rapid replacement)
- Song evolution (gradual change)
- South Pacific song (revolutions sweep west to east)
- North Pacific song (one shared, slowly changing song)
- North Atlantic song
- South Atlantic (Brazil) song
Notes published science
- In a computer model of every Southern Hemisphere humpback population, making all populations the same size sent song revolutions the 'wrong' way (east to west) about half the time (0.54); with the real population sizes, only 5% did. The west-to-east sweep of songs across the South Pacific may be set by which populations are bigger. (src:zandberg-2021)
- When a new song arrives, singers stitch it into the old one at the places where both songs have 'similar sounds arranged in a similar pattern', and they learn songs in chunks (themes), much as songbirds and children do. (src:garland-2017)
- Songs get more complex year by year as they evolve, then drop in complexity every time a revolution brings in a whole new song, a hint that a whale can only learn so much new song at once. (src:allen-2018)
Layers
The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.
humpback whale song: from sound in the sea to a culture across ocean basins (reader's view)
- L1: Sound in sea water: the medium where whales lean on hearing more than on sight or smell; song carries tens of kilometres and more
- L2: Sound makers and singers: mature males' larynges (a U-fold parallel to airflow, a laryngeal sac that lets them sing without exhaling)
- L3: Units: single uninterrupted sounds of up to a few seconds, frequency- or amplitude-modulated, from 20 Hz to upward of 24 kHz
- L4: Phrases and themes: stereotyped arrangements of units, each phrase repeated as a theme; a syntax with periodicities of 6-8 and 180-400 units; language-like statistics
- L5: The song and the session: themes in a fixed order make a 7-30 minute song, repeated for hours by a lone male, his loudness tuned to the wind's noise
- L6: The population song: all males match one song that keeps changing at every level, growing more complex, never returning to old patterns
- L7: The song across populations and in the whales' lives: revolutions that sweep ocean basins west to east, one of a population's three cultural traits, its function in the mating system contested
the boardroom (shared floors)
- L5: The singer's performance: a lone male's long bout, songs repeated without pause for hours, loudness tuned to the noise floor, one voice in an asynchronous chorus; switched on or off by the mating tactic
- L6: The population song and its change: conformity with continual evolution at every level, complexity climbing, novelty entering by copying errors and a bias for novelty, old patterns not revisited
- L6: The song's structure: units in phrases, phrases repeated as themes, themes in a fixed order as a song, suites of themes as song types; a syntax with periodicities that no Markov model captures; language-like statistics
- L7: The song across populations and in the whales' lives: revolutions sweeping ocean basins west to east, one of three cultural traits of a population, and whatever the song does in the mating system (female attraction, male-male sorting, a lek, sonar: contested)
way in: Ambient noise (wind always; vessels, seismic air guns, sonar)
- L1: Wind noise, always present and variable; vessel noise, seismic pulses and fish-imaging sonar added by people
- L2: Masking at the listener: noise shrinks the range a song is heard
- L4: The singer's Lombard response: source level raised with wind noise, not raised further for vessel noise
way in: Echoes of the singer's own song (the contested sonar model)
- L1: Echoes from distant whales or prey and changes in the song's own reverberation
- L2: Song intense enough to make echoes at long range
- L4: Parallel echo streams integrated to follow moving echo sources (the revised 'duplex' model)
way in: Feeding conditions in foraging habitat
- L1: Forage species' availability (krill, anchovy, sardine) set by the ecosystem's state
way in: Neighbouring populations' songs (feeding grounds, migration routes, visits)
- L1: A neighbouring population's song, heard on shared feeding grounds, migration routes or another wintering ground
- L2: Rare contact: in the model a whale learns from a neighbouring population with probability 0.001 on the feeding grounds
- L3: Hybrid songs: whole new themes spliced into the current song, or one hybrid phrase between the two
- L4: Adoption of the whole new song (a revolution), hypothesised to ride on a sexually selected drive for novelty
way in: Other singers' song heard (the current population song)
- L1: Other males' songs in the water: lone singers whose bouts together make an asynchronous chorus
- L2: A listening male within range: song is heard several kilometres from a singer
- L3: Learning in segments: themes and phrase types taken up as units of learning
- L4: Conformity: most males match the current arrangement
way in: Other whales present: joiners, females, competitive groups
- L1: A loose network of interacting individuals and groups on the winter grounds
- L2: A joiner arrives: almost always a lone male not singing at the time
- L4: A conditional mating strategy: switching between singing and non-singing tactics
way in: Season and place: the migration to the winter breeding grounds
- L1: The yearly round: high-latitude summer feeding grounds, low-latitude winter breeding grounds
- L2: The singer: a mature male
way out: Stopping to approach or join another whale
- L1: Movement through the water toward another whale
- L5: A switch of tactic: singers joining females more often under air-gun noise
way out: The song broadcast into the sea
- L1: Sound in sea water at 150-189 dB re 1 µPa at 1 m
- L2: The larynx: a U-shaped fold parallel to airflow, beside a laryngeal sac
- L5: The bout: a song of 7-30 minutes repeated with precision
way out: The song passed on to the next population
- L1: Song heard by other populations' whales on shared feeding grounds and routes
- L7: A population's song type carried eastward to the next population
Sources (19) published science
Garland EC, Corkeron P, Noad MJ et al. 2025. Culture and conservation in baleen whales. Phil Trans R Soc B 380:20240133. doi:10.1098/rstb.2024.0133 (CC BY) (behind its links)
Girola E, Dunlop RA, Noad MJ 2023. Singing humpback whales respond to wind noise, but not to vessel noise. Proc R Soc B 290. doi:10.1098/rspb.2023.0204 (behind its links)
Mercado E 2018. The sonar model for humpback whale song revised. Front Psychol 9:1156. doi:10.3389/fpsyg.2018.01156 (behind its links)
Whale vocalization (Wikipedia, via Wikimedia Enterprise cache; redirected from Whale song) (behind its links)
Payne RS, McVay S 1971. Songs of humpback whales. Science 173:585-597. doi:10.1126/science.173.3997.585 (behind its links)
Zandberg L, Lachlan RF, Lamoni L, Garland EC 2021. Global cultural evolutionary model of humpback whale song. Phil Trans R Soc B 376. doi:10.1098/rstb.2020.0242 (behind its links)
Suzuki R, Buck JR, Tyack PL 2006. Information entropy of humpback whale songs. J Acoust Soc Am 119:1849-66. doi:10.1121/1.2161827 (behind its links)