Reassessing the mechanisms and origins of vocal learning in birds.

The most widely accepted hypothesis of vocal imitation in birds pre-dates many recent studies on the behavior, anatomy, physiology and cell biology of this phenomenon. It states that vocal learning involves two steps: (1) an auditory memory is laid down, and then (2) vocal output is modified until the auditory feedback it generates matches the model. This black-box model of vocal imitation disregards circuitry. We now know that the brain pathways for vocal learning in birds include a series of well-defined nuclei and projections. Some of these nuclei and projections develop late in ontogeny, at the time when auditory models are first acquired and imitated. We also know that the pathways involved in song production respond to sound, an observation that blurs the demarcation between what is an auditory and what is a motor circuit. These and other recent discoveries call for a reassessment of the mechanisms and origins of vocal learning in birds and mammals.