Researchers presented carrion crows with arbitrary visual cues and asked them to produce either a vocalization or a peck. The cues appeared in randomized, alternating or blocked sequences, allowing the researchers to vary how often the crows switched between the two kinds of response.
The crows performed both actions with high accuracy, but vocalizations took longer and varied more in speed. Switching into a vocal response produced a clear delay and occasional errors, whereas switching into pecking had little effect on performance.



