Zebra finches, and the meaning in their calls

One of the oldest questions in biology is where language comes from, and one of its most productive strategies has been to look for its ingredients in other animals: the steps, each one small, that might add up to something language-like. A paper in the 18 September issue of Science reports evidence for a significant step in zebra finches, the small Australian songbirds that have become the standard laboratory subject for vocal learning. The birds, the researchers find, perceive the different types of their own calls in a categorical and semantic way, treating distinct call types as carrying distinct meanings rather than as mere variations of sound.

The distinction being drawn is subtle and important. An animal can react to a sound without understanding it as a category; a bird that responds to a warning call is responding to a particular sound. The paper’s claim goes further: the finches treat their call types as categories with meaning, the way a listener treats words as types carrying reference, rather than as unique acoustic events. The researchers used a paradigm borrowed from studies of human speech perception, habituating the birds to one call type and measuring how they responded when the call type was changed, and the pattern of their responses indicated categorical perception of the calls’ meanings.

The significance of the result is in the architecture it reveals. Categorical perception, treating the continuous variation of sound as discrete meaningful types, is one of the foundations of human language; a spoken word is recognised as a member of a category despite the endless variation of voices and accents. To find that a songbird applies the same trick to its own calls is to find one of language’s deepest habits operating in an animal with no language at all, a finding that helps place the ingredients of speech in the evolutionary inventory long before speech itself.

The study also adds to the growing evidence that the gap between animal communication and human language is not a single wall but a staircase of partial abilities, each one present in some species and absent in others: vocal learning in songbirds and parrots, referential calls in primates and corvids, and now categorical, meaning-like perception in the calls of a small finch. The staircase is the field’s real discovery, and each new step located on it is a step in the map of how language came to be. The finches’ result is the more striking for the modesty of its subject, a bird whose entire vocabulary is a handful of call types, showing that the machinery of categories can operate at the very bottom of the vocal ladder, wherever there is a sound that means something to its maker.

The finches themselves, unaware of their new prominence, will go on calling as they always have, their small categories and meanings intact. But the paper has added their calls to the evidence on which the story of language is built, and the story’s shape is growing clearer: the abilities that language is made of did not appear all at once with speech; they accumulated, piece by piece, in the long evolutionary run-up, and zebra finches, it turns out, have been practising one of them, in their tiny categorical way, all along.