Songbirds can be trained to understand non-language aspects of human speech, a study has found.

Scientists showed that zebra finches are even more sensitive to “prosodic” variations, such as changes in the pitch, volume and duration of spoken syllables, than humans.

The findings suggest that prosodic features of vocal sounds may have pre-dated the evolution of structured language.

For the study, birds were trained to respond to patterns of human speech syllables played through a loudspeaker above their cage.

Depending on the sound, the birds had to peck a specific key or take no action. A correct response was rewarded with access to food, while an incorrect one led to 15 seconds of darkness.

Two different sequences of four naturally-spoken syllables were uttered by both a male and female speaker.

Prosodic patterns typical of human speech were added by altering pitch, volume or duration to stress the first or last syllable in each group.

The finches not only recognised the changing patterns, but applied what they learned to sequences consisting of new syllables.

“Zebra finches are sensitive to the same prosodic cues known to affect human speech perception,” the scientists, led by Michelle Spierings, from Leiden University in the Netherlands, wrote in the journal Proceedings of the Royal Society B.

They added: “Our results show that sensitivity to prosodic cues is not linked to the possession of language and might have preceded language evolution, poss-ibly originating from a pre-existing sensitivity to meaningful variation in pre-linguistic communicative sounds.”