A simple computational principle predicts vocal adaptation dynamics across age and error size.

Area of Science:

• Neuroscience
• Bioacoustics
• Animal Behavior

Background:

• Vocal learning in songbirds and humans relies on sensory feedback for error correction.
• Acoustic variability decreases with age, impacting vocal error correction.
• Previous research indicates adults correct small errors but not large ones.

Purpose of the Study:

• To investigate age-dependent changes in vocal error correction speed and magnitude.
• To test if younger animals correct large auditory errors more readily than older ones.
• To determine if a common computational principle governs sensory-driven motor learning across ages.

Main Methods:

• Experimental manipulation of auditory feedback in songbirds of different ages.
• Quantification of vocal changes in response to auditory errors.
• Modeling the relationship between sensory error, prior feedback, and learning dynamics.

Main Results:

• Vocal error correction speed and extent in songbirds change significantly with age.
• Age-dependent learning differences are accurately predicted by a model.
• The model incorporates the overlap between sensory errors and prior sensory feedback distribution.

Conclusions:

• The brain uses a consistent computational principle to adjust vocal adaptation.
• This principle calibrates learning rate and magnitude based on age and error size.
• Findings offer insights into the neural basis of vocal learning and adaptation.