
The question
Neuroscientists commonly learn how the brain processes communication by playing recorded vocalizations to an animal. But hearing someone’s voice played through a speaker isn’t the same thing as actually having a conversation with them.
Does the brain treat them as equivalent?
How we did it
We recorded the same individual neurons in marmoset prefrontal cortex while the animals heard the same type of vocal signal under three conditions: restrained, freely moving, and freely moving while participating in an interactive vocal exchange.
What we found
We found that the same type of marmoset vocalization could produce very different responses in prefrontal cortex depending on what the animal was doing when it heard the sound. A neuron that responded strongly to a call in one behavioral context might respond weakly, or not at all, to that same kind of call in another context. Consequently, knowing how a neuron responded to a vocalization under one condition was a poor predictor of how it would respond when the animal was engaged in a different behavior.
These differences were not limited to individual neurons. Across the population, patterns of neural activity distinguished the behavioral context with greater than 90% accuracy. The responses observed during natural vocal communication were particularly distinct from those recorded when the animals were simply listening to sounds. We also found that neural activity in prefrontal cortex before another marmoset called helped predict whether the animal would subsequently respond. This suggests that the brain’s ongoing state contributes to how an incoming vocal signal is processed during a social interaction.
These findings demonstrate that the neural representation of communication cannot be understood simply by measuring responses to isolated sounds. In natural interactions, the brain must integrate incoming vocal signals with the animal’s ongoing behavior, the social situation, and the opportunity to respond. The same acoustic signal can therefore be processed differently depending on the circumstances in which it occurs.
Why it matters
Many neurological and neuropsychiatric disorders affect social communication, including autism spectrum disorder, schizophrenia, and brain injuries that disrupt frontal systems. The clinical difficulty is that communication problems often emerge most clearly during real interactions—when a person must interpret who is speaking, what has already happened, and when and how to respond—not when isolated sounds are presented in a laboratory test.
By showing that the same vocal signal can be represented differently depending on the behavioral context, this work provides a framework for developing more realistic ways to study social-communication dysfunction. Ultimately, understanding the neural mechanisms that support interactive communication may help identify which components of this system are altered in disease and provide more behaviorally meaningful measures for evaluating dysfunction and recovery.