Bonobo Communication: Study Reveals Human-Like Language Compositionality

A new study published in *Science* on April 4, 2025, reveals that bonobos (Pan paniscus) exhibit compositionality, a linguistic feature previously considered unique to humans. Researchers from the University of Zurich and Harvard University analyzed hundreds of hours of recordings from a reserve in the Democratic Republic of Congo. Compositionality is the ability to combine linguistic elements into larger, more complex structures to generate diverse meanings. The study identified non-trivial compositionality in bonobo vocalizations, where the meaning of the whole is formed by interdependent parts. This contrasts with trivial compositionality, where elements are independent. Researchers, including Mélissa Berthet, Simon Townsend, and Martin Surbeck, recorded and analyzed over 700 bonobo vocalizations, noting the context of each vocalization, including activities, location, companions, and nearby elements. This allowed them to create a "dictionary" of bonobo sounds. They then used a mathematical method from human linguistics to analyze how frequently different sounds appeared together. The analysis revealed compositionality in at least four combinations. For example, a high-hoot (“pay attention to me”) combined with a low-hoot (“I am excited”) together appeared to mean something like: “Pay attention to me, because I am in distress.” And a peep (“I would like to…”) with a whistle (“let's stay together”) was used in intimate contexts like sex or displays of dominance. This finding suggests that compositionality may have emerged earlier in primate evolution than previously thought. Humans and bonobos share a common ancestor from approximately 7 to 13 million years ago, suggesting that the cognitive building blocks that facilitate this capacity are at least seven million years old.