How chimpanzee communication reflects the same language structures as human communication

Chimpanzees Asanti and Akuna vocalising. Credit: Liran Samuni / Taï Chimpanzee Project

Chimpanzees are not just limited to emitting isolated verses, but can combine them with a complexity that echoes human language structures: this was documented in a peer-reviewed study conducted by an international team of French and German researchers and published in the scientific journal Science Advances. Analyzing thousands of vocalizations from three groups of chimpanzees in Taï National Park, Côte d'Ivoire, the authors detailed how these animals know how to modify or generate different meanings through different combinations of two calls, made according to four distinct patterns. It is a system that experts say could even trace back to the common ancestor between humans and great apes and challenges the idea that advanced linguistic ability is exclusive to our species. “The complexity of this system suggests either that there is indeed something special about hominid communication or that we have underestimated the complexity of communication in other animals as well, which requires further study,” the authors comment.

Syntax, or the set of rules that determine how sounds combine to produce meanings, is considered one of the central components of human language. Until now, in fact, other species were thought to rely on a limited number of isolated calls, with few combinations mostly reserved for alarm contexts, such as the presence of predators. It is thus a view that, for decades, supported the idea that the communication systems of nonhuman primates were too rudimentary to be considered precursors of language. However, it is safe to speculate that such assumptions will soon be reevaluated, as the newly published study suggests that great ape communication may be much more sophisticated than previously thought. According to the authors, primates would use structures that mirror known principles of human syntax and semantics, which would open up new scenarios about the evolution of our ability to speak.

Specifically, researchers from the Max Planck Institutes in Leipzig, together with colleagues from the CNRS centers and Université Claude Bernard Lyon 1, recorded thousands of vocalizations of three groups of wild chimpanzees in their natural habitat, examining twelve distinct types of calls and observing how the meaning changed when they were combined two by two. According to their results, some combinations were “compositional”-that is, where meaning resulted from the sum of the two elements (such as a call for ‘feeding’ followed by one for “rest,” indicating a situation involving both behaviors)-while others clarified or specified the meaning of the first call. Finally, the most surprising, namely the so-called non-compositional idiomatic combinations, in which the sequence produced an entirely new meaning that could not be traced back to the original calls.

“The chimpanzees also used non-compositional idiomatic combinations that created entirely new meanings (e.g., A = rest, B = affiliation, AB = nesting),” the researchers in fact explain. "Our results suggest a highly generative vocal communication system that is unprecedented in the animal kingdom, echoing recent findings on bonobos that complex combinatorial abilities were already present in the common ancestor of humans and these two great ape species. This changes the view of the last century, which viewed communication in great apes as fixed and tied to emotional states, and thus unable to provide us with information about the evolution of language. Instead, here we see clear indications that most types of recall in the repertoire can change or combine their meaning when combined with other types of recall," concludes co-author Cédric Girard-Buttoz, adding that further studies will be needed to investigate whether there is something special about hominid communication, whether we have underestimated the complexity of animal communication, or whether perhaps both hypotheses are true.

https://www.science.org/doi/10.1126/sciadv.adq2879