Music is often referred to as a “universal language,” but is that claim really accurate? A study published in Current Biology on Thursday, Jan. 25 suggests that the answer is yes.
Researchers from Harvard University ran a study with two experiments to test whether the topic of a song from one culture could be identified by people from a completely different culture. For the study, the researchers chose to focus on songs including human singing, not just instruments. Their experiments were based on the idea of form-function relations, which states that certain forms of sounds and vocalizations are reliably linked to certain functions or uses in many animals and in humans. For example, it has been established in many species of vertebrates that low-frequency, harsh vocal sounds function to convey hostility.
In the researchers’ first experiment, 750 people from 60 different countries listened via the internet to very brief snippets of 36 songs randomly selected from a group of 118. The songs came from 86 small-scale societies, including groups of hunter-gatherers, pastoralists and subsistence farmers, and represented 75 languages. Each song chosen was considered part of one of four groups: dance song, love song, healing song or lullaby. These categories were used because they are present in most cultures and have been shown to have connections to the biological and cultural evolution of music.
Participants listened to each snippet and rated the degree to which they thought the song was intended “for dancing,” “to soothe a baby,” “to heal illness,” “to express love for another person,” “to mourn the dead” and “to tell a story.” The researchers found that listeners correctly identified dance songs and lullabies the vast majority of the time, despite being unfamiliar with the songs and the cultures they came from. Listeners were also able to identify healing songs at a slightly lower rate of accuracy. Listeners consistently assigning the correct function to unfamiliar songs in languages different from their own supports the idea that universal form-function relations exist in human song. In other words, the experiment supported the idea that songs fulfilling a certain function, such as lulling a child to sleep, tend to have a certain set of features regardless of the culture producing them.
To examine which features each group of songs had in common that enabled people to recognize them, the researchers ran a second experiment. They had 1,000 people, half from the United States and half from India, listen via the internet to very brief snippets of 18 randomly selected songs from the same group of 118 songs. Then, instead of asking people to identify the function of the songs, they asked them to report each song’s number of singers, gender of singer(s), number of instruments, melodic complexity, rhythmic complexity, tempo, steady beat, arousal, valence and pleasantness.
When the researchers analyzed all the ratings, they found dance songs were consistently reported to have more singers, more instruments, higher melodic complexity, higher rhythmic complexity, faster tempo, steadier beat, higher arousal, higher valence and higher pleasantness. Additionally, lullabies were consistently reported as being low in all of the categories in which dance songs were high, and they were also very likely to have a female singer. The results for healing songs and love songs were inconclusive. Based on their analysis, researchers believe there are additional shared features for each group of songs that they did not identify in their second experiment, and these could be explored in future studies.
Overall, the researchers’ results support something people have suspected throughout history: songs, in one way or another, are a part of us all.
Abbey Bigler is a fourth-year student majoring in English with minors in business and technical writing, communications studies and biology. ✉️ AB842693@wcupa.edu.