Suppose you play recordings of songs from societies around the world for listeners who have never heard them. Some of the songs were made to put a baby to sleep. Others were made for dancing. Can the listeners tell which is which? In a widely discussed set of studies from the Natural History of Song project, they often could, especially for lullabies and dance songs. That result is part of a larger, more careful story about what is and is not shared in human music.
Myth: music is a universal language
The phrase “music is a universal language” is appealing and misleading. Music is not a language with a shared vocabulary; a melody that signals mourning in one place might accompany a celebration in another. The better question is narrower: are there recurring patterns in how people make and use song?
The tidy claim
“Everyone hears music the same way, so music must be a universal language that needs no learning.”
What the research supports
Music of some kind appears in every society studied, certain song uses recur, and listeners can sometimes guess a song’s purpose. But taste, scales, and what counts as pleasant vary, and variation inside a society is large.
What Mehr and colleagues found
In 2019, Samuel Mehr and colleagues published “Universality and diversity in human song” in Science. According to a summary on the US National Institutes of Health director’s blog, the team drew on the Human Relations Area Files, a cross-cultural ethnographic collection covering 319 societies, and found music mentioned for 309, with other records indicating the remaining six also made music. They also analyzed close to 5,000 vocal songs from 60 societies across all continents.
Three findings stand out. Song appears in similar contexts around the world, including dance, love, healing, and caring for infants. Songs in those contexts share some features, so acoustic properties help predict what a song is used for. And the way song varies can be described along three broad dimensions the authors call formality, arousal, and religiosity. A public online experiment with nearly 30,000 participants in over 100 countries found that listeners could often tell what a song was for, even from far away.
The same summary notes that variation within a society was about six times greater than variation across societies. That is worth remembering: the study documents both shared tendencies and wide diversity, and it does not suggest every culture’s music is alike.
Earlier listening experiments
A 2018 study in Current Biology asked 750 English-speaking online listeners from 60 countries to judge the purpose of songs from small-scale societies. Listeners were most confident about lullabies and dance songs and less so about healing songs; love songs were hard to identify. Lullabies were heard as slower and simpler, dance songs as faster and more complex. The caveats are real: all the participants in the first experiment spoke English, a second experiment drew on listeners in the United States and India, and an outside expert quoted by Smithsonian magazine cautioned that, by itself, the result says little about universals.
Where preferences differ
Even sounds that feel natural to some listeners are not heard that way by everyone. In a 2016 study in Nature, researchers working with more than 100 Tsimane’ people in the Bolivian Amazon, who have limited exposure to Western music, found that the Tsimane’ rated consonant and dissonant sounds as roughly equally pleasing, while comparison groups with more exposure to Western music preferred consonance. The researchers concluded that culture plays a large role in shaping musical taste. A fair note of caution: the groups differ in many ways besides music exposure, so the comparison cannot isolate a single cause.
Why did music arise? Competing hypotheses
This is the part where careful writing matters most. Music leaves no fossils, and no one knows when or why it began. Researchers have proposed several ideas:
- Social bonding. Making music together may help groups cohere.
- Credible signaling. Mehr, Krasnow, Bryant, and Hagen argued in Behavioral and Brain Sciences that music works as a hard-to-fake signal, for example by a coalition showing its strength or a caregiver showing attention to an infant.
- A by-product. Some argue music piggybacks on other abilities; a catchy nickname for this view is “auditory cheesecake.” Even one author of the signaling paper has said the idea cannot be entirely dismissed.
In a reply to that paper, Patrick Savage and colleagues argued that bonding and signaling are complementary rather than rival explanations, and that a gene-and-culture view fits better than a strict adaptation-versus-by-product split. They also acknowledge that hard evidence for the earliest stages is absent and that a well-known Neanderthal-site “flute” may have been chewed by carnivores.
Why it matters in everyday life
Whatever its origins, music shows up in weddings, work, worship, protest, and comfort. If you want to see how music sits alongside other shared forms of expression, read about why humans tell stories or the human urge to create. And for what group rituals may do for belonging, see why humans need to belong.
