Summary: New research shows that learning and retaining the facial and vocal movements required for speech depends primarily on the brain’s sensory systems—auditory and somatosensory cortices—rather than on the primary motor cortex. This finding reshapes theories of speech motor learning and suggests new directions for speech-restoration technologies and stroke rehabilitation.
Researchers found that temporarily disrupting neural activity in the auditory or somatosensory cortex severely impaired participants’ ability to retain newly learned changes in speech production. By contrast, disrupting the primary motor cortex had no measurable effect on retention. These results point to sensory plasticity as the key substrate for forming and storing speech motor memories.
Key findings
- Sensory dominance: Contrary to longstanding assumptions that frontal motor areas were the principal locus of speech learning and memory, this study demonstrates that sensory cortex plasticity drives retention of newly learned speech movements.
- Real-time altered feedback: Participants learned new speech patterns while researchers altered the acoustics of their vocalizations in real time and played the modified sound back through headphones, prompting rapid, automatic speech adjustments.
- TMS retention test: After learning, investigators used non-invasive transcranial magnetic stimulation (TMS) to transiently disrupt either the auditory cortex, the posterior somatosensory cortex, or the primary motor cortex. Memory for the learned speech pattern was tested 24 hours later.
- Sensory disruption impairs memory: Disrupting the auditory or somatosensory cortex before the retention test strongly reduced participants’ ability to reproduce the learned speech changes. Disrupting the motor cortex did not impair retention.
- General principle of sensorimotor plasticity: The results align with related work on limb motor learning showing that sensory cortex plasticity supports retention of learned movements, suggesting a broad role for sensory systems in motor memory.
- Implications for technology and rehabilitation: A sensory-first architecture for brain–speech interfaces and rehabilitation devices could improve performance and usability, especially in post-stroke speech recovery where restoring sensory feedback networks may accelerate relearning.
Source: McGill University
Overview
This collaborative study from McGill University and the Yale School of Medicine tested whether memory for speech movements is stored in motor areas or in sensory cortices that process sound and touch. The team used a well-controlled laboratory model in which altered auditory feedback drives rapid speech motor learning, a method that engages auditory, somatosensory, and motor systems.
Retention tested through brain stimulation
Participants first learned a new speech pattern while hearing a modified version of their own voice through headphones. This induced compensatory changes in articulation and vocal output. After the learning phase, researchers applied targeted TMS to temporarily disrupt activity in one of three cortical regions: the superior temporal gyrus (auditory cortex, STG), the posterior somatosensory cortex (S1), or the primary motor cortex (M1).
Retention of the learned speech movement was assessed 24 hours later. The hypothesis was simple: if a region is necessary for forming and maintaining speech motor memory, disrupting it would impair retention. The results were clear—disruption of either STG or S1 degraded retention, while disruption of M1 produced performance indistinguishable from a no-TMS control.
Importantly, the disruption effects were specific to the memory for newly learned speech patterns; basic speech production remained intact. The data therefore support the view that sensory cortical plasticity underlies memory for speech movements, enabling the production of newly learned articulatory patterns.
The role of brain plasticity
This work is part of a broader research program examining how plasticity in sensory cortices supports motor learning and retention across modalities. Prior studies from the same group on upper-limb movements produced parallel findings: interfering with sensory cortex activity impairs the retention of newly learned motor skills. Mapping the cortical circuits that link sensory representations to motor output is a priority for future research.
Translational efforts will explore sensory-targeted interventions to enhance rehabilitation after neurological injury. For stroke survivors and others with impaired speech, therapies and devices that restore or augment sensory feedback—rather than focusing solely on decoding motor intentions—may better support the brain’s natural mechanisms for relearning speech.
Funding
This research was supported by the U.S. National Institute on Deafness and Other Communication Disorders.
Frequently asked questions
A: Learning to speak relies on matching outcomes to sensory expectations—how sounds and articulatory sensations should feel and sound. Sensory cortices contain the representations that encode these expected outcomes. The motor cortex executes movements but the sensory systems store the templates used to guide and recall those movements.
A: The experiment induced learning with altered auditory feedback played through headphones, provoking automatic adjustments in speech. After learning, researchers used TMS to transiently disrupt different cortical regions. Participants whose sensory cortices were disrupted showed impaired retention of the learned speech changes; those with disrupted motor cortex did not.
A: Many current devices emphasize decoding motor intent. These findings suggest therapies and interfaces that restore or enhance sensory feedback—auditory and somatosensory—could better support the brain’s own learning mechanisms and improve outcomes for people relearning speech after injury.
Editorial notes
- This article was edited by a Neuroscience News editor.
- The original journal paper was reviewed in full.
- Additional context was added by editorial staff.
About this research
Author: Kay Pettigrew
Source: McGill University
Contact: Kay Pettigrew – McGill University
Image: The image is credited to Neuroscience News
Original research: Open access. “Sensory Basis of Speech Motor Learning and Memory” by Nishant Rao, Rosalie Gendron, Timothy F. Manning, and David J. Ostry. PNAS. DOI: 10.1073/pnas.2525468123
Abstract
Sensory Basis of Speech Motor Learning and Memory
Speech motor learning using altered auditory feedback provides a model that engages auditory, somatosensory, and motor components. In this study, transcranial magnetic stimulation was used to disrupt the superior temporal gyrus (STG), posterior somatosensory cortex (S1), or primary motor cortex (M1) after speech motor learning. Retention was tested 24 hours later. Disruption of STG or S1 impaired retention of the learned speech changes, while disruption of M1 did not differ from a no-TMS control. The effects were specific to memory for learned speech patterns and did not interfere with basic speech production. These results support the conclusion that plasticity in auditory and somatosensory cortices is necessary for speech motor learning and memory.