Vocal mimicry in songbirds

Baylis (1982, Acoustic Communication in Birds, Academic Press) decried the serious lack of experimental verification for the various hypotheses proposed to explain vocal mimicry in songbirds. With few exceptions, our understanding of the function and acquisition of this fascinating behaviour seems to have scarcely progressed. We examine the proposed functional explanations and supporting evidence, and summarize advances made since Baylis's (1982) review. We conclude that there is no compelling evidence to support any of the functional hypotheses but, rather, that almost all of the data concerning song mimicry are consistent with the learning mistakes hypothesis, whereby birds learn simple and common sounds, frequently using them in inappropriate contexts. Additionally, many apparently mimicked sounds are calls, not songs, which themselves may not be learned by the models. It is plausible that many examples of call mimicry are, in fact, due to evolutionary convergence.