Vocal comprehension learning is widespread across birds

For decades, scientists have known that only a few groups of birds—songbirds, parrots, and hummingbirds—can learn to produce new sounds. But a new article in The Quarterly Review of Biology, Volume 100, Number 4, reveals that many more birds can learn to understand the sounds of others, suggesting that comprehension learning, not production, may be the foundation for the evolution of language.

In their paper, “Vocal Contextual Learning in Birds,” William A. Searcy and Eleanor M. Hay of the University of Miami review decades of research to uncover how birds modify their understanding and use of vocal signals through experience. Their findings show that vocal comprehension learning occurs naturally in at least 17 of 37 avian orders, encompassing nearly every major branch of the bird family tree.

“The ability to extract information from sounds is better developed in birds than is their ability to encode information in sound,” the authors write.

Drawing on hundreds of studies, the authors identify striking examples of vocal comprehension learning in nature: gulls that recognize mates by their calls, penguin parents that find their chicks in crowded colonies through individual “signature calls,” and songbirds that eavesdrop on the aggressive or courtship interactions of neighbors to infer social relationships. Some species, like the cooperatively breeding acorn woodpecker, even show “third-party knowledge,” recognizing social relationships they’re not directly part of.

In contrast, vocal usage learning, in which an individual modifies the circumstances in which it produces specific vocalizations, appears in only four orders, and often only under laboratory training. Most cases come from conditioning studies in the lab, such as African grey parrots trained to label colors, shapes, and numbers. Natural examples include canebrake wrens that learn new duet patterns when they acquire new mates, or zebra finches that adjust the timing of their calls to avoid overlapping others’ songs.

The paper’s evolutionary analysis shows that comprehension learning likely originated early in bird evolution while usage learning appears to have evolved later and more sporadically. The researchers also suggest that comprehension learning may confer broader evolutionary benefits. Recognizing a mate, offspring, or a neighbor’s song can directly affect survival and reproduction.

By revealing that birds are far better at understanding vocal signals than inventing new ways to use them, the authors offer a perspective on how language may have evolved in humans. Language, they argue, may have arisen first through expanding the ability to comprehend meanings in sounds, with complex usage, and eventually syntax, developing later.

“The most striking pattern emerging from this review of vocal contextual learning in birds is the rarity of instances of vocal usage learning,” the authors conclude.  “Even referential signaling systems in birds, which often have just two signal categories, may be simple enough that learning is not needed to produce correct usage.” They note that it is in vocal behaviors with more complex usage patterns, such as in duetting using duet codes or in referential systems of greater complexity—if they exist in birds—that further evidence for vocal usage learning should be sought.

The premier review journal in biology, The Quarterly Review of Biology has presented insightful historical, philosophical, and technical treatments of important biological topics since 1926. The QRB publishes outstanding review articles of generous length that are guided by an expansive, inclusive, and often humanistic understanding of biology.

Contact: Mallory Gevaert / mgevaert@uchicago.edu