To human ears, birdsong is poetic, wholesome and romantic—even transcendental. But as far as most birds are concerned, “it’s all sex and violence”, says Kristal Cain, an evolutionary ecologist at the University of Auckland. While some birds woo each other with operatic duets, she says, others use song in posturing rap battles with rivals or to sabotage their mate’s flings with other birds.
From Darwin on, ornithological researchers—the vast majority of whom, in Cain’s words, were “tweed-wearing, whiskered, old white field biologists”—defined birdsong as something male birds did in the breeding season to mark their territory and attract mates. They also assumed that the behaviour common to migratory European and North American birds was standard across the entire avian world. Female birds rarely sang, they declared—and if they did, it was likely the result of “hormonal aberrations”.
However, as more women became biologists, and as people began looking more closely at birds from the tropics and the southern hemisphere, it became apparent that this assumption was completely backwards. Females sing in the majority of songbird species, and the ancestor of all of them—the mother songbird—also sang.
“It’s actually the norm,” says Cain, “and now we know that, when song evolved, it evolved in males and females.” Those northern hemisphere female birds just lost the ability somewhere along their evolutionary journey.
A 2020 study of academic papers about female birdsong over the past 20 years found that 68 per cent were written by women. “They’re asking completely different questions than the traditional dudes would ask,” says Cain, “and they’re finding evidence for things that have never been looked at before.”
To compare the purpose of male and female song, Cain and her colleagues recorded the songs of nine enthusiastically named species of Australian fairy wrens—the family includes the splendid fairy wren, the superb fairy wren and the lovely fairy wren. Presumably, it was the males’ jewel-like plumage that so impressed the naming naturalists, because the dun-coloured females look like “animated balls of dust”, says Cain.
But looks aren’t everything. Cain’s study revealed that female fairy wrens sing songs that are just as complex as the males’ melodies. They chirp away right through the breeding season, whether they’re fertile or not. So, if they’re not wooing, what are they singing for? It’s mainly to defend their nests and territories. “Females that sing more frequently tend to have more successful nests,” says Cain. “Some territories have better access to food and water, or fewer predators, and females have just as much of an incentive to fight over those things as males do.”
Fighting songs are common in many songbirds: a female korimako, or bellbird, will fly aggressively at another female who enters her territory, get up in the stranger’s face, and sing. These songs function as theatre and threat—and if diplomacy fails, things can get bloody. In some species, female birds will fight to the death over the best nest cavities.
Other birds use song for a different purpose: to cock-block their partners. Female Peruvian warbling antbirds chatter over top of their mate to drown him out—biologists call it signal jamming—and prevent him from flirting with another bird. “Like, ‘You don’t sing to her! You’re busy over here, buddy! You’ve got a nest you need to be taking care of!’” says Cain. “And she’s telling that other female to stay the f— away.”
Birds are far from the only creatures making music. Bowhead whales shriek, moan, cry, rattle, whistle and hum. Gibbons perform arias with their mates, using vocal mechanisms common in human opera singers. In central America, singing mice duet at ultrasonic frequencies too high for us to hear. In New Zealand, male short-tailed bats croon to the ladies for six hours a night, combining four notes into dozens of syllables while spraying themselves with urine.
But what actually counts as a song in the animal world is a matter of debate. Is a blue whale’s subsonic monotone a song? How about a kākāpō’s boom, or a leopard seal’s glissando moan? Does it need to have a tune to be a song? Or is it what the animal is trying to say that matters? Different branches of biology define song differently—and while humans like to categorise everything into tidy boxes, “biology doesn’t give a shit what we want”, says Cain.
In ornithology, researchers distinguish birds’ songs from their calls (or try to), and the difference relates to complexity and purpose. Calls tend to be shorter, simpler and directed at friends. “‘I’m flying over here, I found some food there, I’m bringing you some food, there’s an owl over there’—those types of things,” says Cain. Songs are generally louder, more complicated, and broadcast across the whole neighbourhood.
Most birds are born already knowing their species’ songs, but others must learn how to sing from the adults around them. These birds, called vocal learners, can only babble at first, like infant humans learning to speak.
They must spend hours listening, practising (and getting it wrong) until their adult song crystallises. Cain and her colleagues looked at how long it takes different species to learn their grown-up songs, and “it’s all over the damn map”. Zebra finches are pitch-perfect just 50 days into their existence. Korimako take about six months, while other birds need to migrate and return the following year before they sound like adults.
A handful of species at the upper end of this scale don’t stop learning when they mature, and continue to learn new calls and songs throughout their lives. Australian lyrebirds, for example, can mimic almost anything. “A tūī can develop a car alarm, and the next year sound like a squeaky bike tyre, and the next year sound like a fantail,” says Cain.
Orcas have learned the dialects of the dolphins they hang around with. And a 2022 study by Jenny Allen from the University of Queensland found that male humpback whales from New Caledonia could learn incredibly complex songs from East Australian humpbacks while the two species were feeding together in the Southern Ocean—and then reproduce the songs, note-perfect, when they returned home. “It’s rare for this degree of cultural exchange to be documented on such a large scale in a non-human species,” Allen says.
It helps to think of birdsong more as language than music, says Dianne Brunton, a professor of behavioural ecology at Massey University. Songs are made up of repeated syllables or elements, “a bit like words in a sentence”.
Vocal learning opens up the possibility of cultural evolution, says Brunton, as animals improvise or make mistakes, then pass on altered songs. She has spent many years studying tīeke, or North Island saddlebacks, which have become famous for their regional dialects.
Driven almost to extinction by mammalian predators, by the early 1900s tīeke survived only on Taranga/Hen Island, in the northern Hauraki Gulf. A series of translocations starting in the 1960s—designed to make sure the tīeke weren’t laying all their eggs in one vulnerable basket—gave researchers the chance to study how much their dialects changed over time. Brunton’s student Kevin Parker made 2700 recordings of tīeke on 13 islands, identifying 202 different songs. Only 30 per cent of those songs were shared between islands, and when Parker played unfamiliar songs to the birds on Motuihe, he got no response. What if the birds’ songs were now so different they no longer recognised each other as tīeke?
A subsequent study in Auckland’s Shakespear Regional Park proved that love can cross the language barrier—with a tragic ending that would have moved the Bard himself. Parker moved tīeke who sang in three different dialects into the park. Initially, most birds got together with their own kind—the ones from the same island that sang in the same dialect. But, like immigrant kids the world over, their children had different ideas. Young birds born at Shakespear paired up with those from different backgrounds, and the collision resulted in a cultural flowering. A subsequent study by another student, Kyle Sutherland, found five new types of song in that first generation.
The young tīeke were coming up with their own slang. “Actually, memes,” says Brunton. Yes, memes. Before the word was hijacked by the internet, it was coined by Richard Dawkins to mean an idea spread through a population via cultural evolution, in the same way that a gene is spread through genetic evolution. In the tīeke context, the meme is the repeated phrase in the middle part of a bird’s song. The Shakespear study demonstrated the flexibility of tīeke songs, and how quickly innovation can happen.
But in 2020, a stoat sneaked through the predator-proof fence protecting the park and had babies. The stoat family mowed down the thriving population of tīeke, leaving just half a dozen pairs—and extinguishing the new songs. “We showed that you could maximise genetic diversity and cultural diversity,” says Brunton, “but then ratbag stoats came along, and that was it.”
Next, Brunton plans to study how that cultural evolution happens in tīeke, korimako and tūī. Do birds prefer to sing common songs or rare ones? Or, instead, do they copy the songs of high-status individuals? In Brunton’s words: “Are there influencers in these social networks?”
In Australia, regent honeyeaters are at risk of losing their love songs entirely—a romantic calamity that could send the species hurtling towards extinction.
Only around 200 to 400 birds remain, making the regent honeyeater Australia’s most threatened songbird, and as rare as kākāpō. Once, they roamed throughout southeastern Australia in flocks of hundreds, but their favourite habitat—the kind of eucalyptus forest that grows in good soils—has been systematically toppled for agriculture since Australia’s colonisation. The remaining birds live across a vast area of New South Wales, from the Blue Mountains to the Northern Tablelands.
Researchers occasionally get the chance to study honeyeaters in the wild when they turn up in one of their semi-regular breeding spots, says the Australian National University’s Ross Crates. “But apart from that window, they sort of just disappear into the landscape, and we don’t really know where they go. None of us really know exactly how they live their lives.”
The birds’ mysterious peregrinations make them a hard species to study—which is why the honeyeaters are a matter for the ANU’s Difficult Bird Research Group. (Other birds studied by the group are similarly mobile, elusive and extremely endangered.)
Honeyeaters are vocal learners, but males do not sing while they rear their chicks, and parents kick the young out of the nest and chase them from their territory when they’re only eight weeks old—before they begin to sing. Teenage males must learn their wooing songs from other males. But there are so few honeyeaters that often the teenagers can’t find any mentors to teach them.
Crates and his colleagues found that 27 per cent of males sang funky songs that differed from the regional norm. A tragic 12 per cent failed to sing any honeyeater songs at all, instead copying the tunes of other species.
“It’s funny when you hear a regent singing like a noisy friarbird, but it’s also heartbreaking,” says Crates. The males that didn’t know honeyeater love songs were, unsurprisingly, less likely to find a mate.
The study sheds light on the losses that come with extinction, even before an animal is technically extinct. Once culture is lost, can a species come back? Is a regent honeyeater still as much of a regent honeyeater if it can’t sing the right song?
The Difficult Bird Research Group is trying to help. At Taronga Zoo in Sydney, researchers are making regent honeyeater mixtapes and playing them to juveniles, who are housed in the same enclosure as a few mature males to act as tutors.
“The idea is that if we can get them to sing properly, then in future years, those birds will get sent back out into the wild,” Crates says. “The next five years are going to be really crucial.”
If a male regent honeyeater sings the wrong song, he might not get laid, but the stakes are even higher for male Saint Andrew’s Cross spiders.
In fairy wrens, males get all the fun colouring, but it’s the reverse for these Australian arachnids. The male is small and brown, with a body just a few millimetres across. The female is striped silver, red, black and yellow, and she’s a giant in comparison—with her legs outstretched in their four-point cross, she would easily fill your palm. And, like most web-building spiders, she’s also short-sighted. “So, when a male comes to a female in her web, he’s really taking his life into his own hands,” says behavioural ecologist Anne Wignall from Massey University. “If he steps on to the web and creates vibrations, then the female might go, ‘Ooh, food!’”
To avoid being eaten, the male rocks back and forward, oscillating the silk with his body in a rhythmic shudder, like playing a single-stringed guitar. The female senses the vibrations with receptors located in special slits on her legs and, if the male can withstand the life-or-death performance pressure and do it just right, his song will calm her down long enough for him to tightrope-walk towards her and attempt to mate.
Wignall’s team recorded spiders’ shudders by pointing a laser beam at the bouncing strands of silk and measuring how the light shifted as the strands vibrated. Different species of web-building spider performed different rhythms, they found. But without technology, these spider songs aren’t audible (or visible) to human senses. So is it really singing? “It’s serving the same function as birdsong,” says Wignall. “It’s just being produced in a slightly different way by an animal that’s a little bit smaller. And with more legs.”
For humans, there’s something special about singing. It’s not just communication—the human voice raised in song has an emotional power that’s unmatched in the animal world (as far as we know). Songs have the ability to move us to tears, conjure up a long-lost lover, and fill us with nationalistic or religious fervour. We sing to express our feelings and to tell stories. “Songs are really quite unique to humans,” says Dianne Brunton.
Still, it seems there’s something about the idea of animals singing that makes them more relatable. Released in 1970 by bioacoustics expert Roger Payne, Songs of the Humpback Whale sold 125,000 copies and became the most popular nature sounds album of all time. NASA included one track on the Golden Record it launched into space in 1977. Pete Seeger and Judy Collins wrote songs inspired by it, and, as Payne had hoped, it helped spur the anti-whaling movement.
Could other animal songs foster a better connection with the natural world? If regent honeyeaters lose their love songs, will that move us to help them? Now that we know spiders sing, might we treat them differently?
“People don’t really think about what invertebrates might be saying to each other,” says Wignall. “But once you really look at them, they’ve got all these complex, amazing behaviours. There are so many similarities, even though we look so different.”