Fluent speech is guided by a sensorimotor error-detection network in the brain

Research by Adeen Flinker and Müge Özker at New York University reports regions of the brain that signal when the speech we try to produce doesn’t match what we hear. Publishing February 3^rd in the open access journal PLOS Biology, the study shows that hearing one’s own speech with unnatural delays leads to matching increases in brain activity related to both hearing and making voluntary movements. These error signals allow us to automatically make corrections and change how we speak, keeping speech smooth and fluent.

Being unable to speak fluently, such as having a stutter, can often be traced back to faulty auditory feedback control—an inability to hear one’s own voice and automatically correct errors. While behavioral therapy can help, knowing where in the brain the error signal is generated is important for developing other methods of treatment, such as biofeedback control.

To identify the error signal, the researchers recorded EEG activity in the brain while people read sentences and heard their voice on delay, much like the echo of a poor teleconference. However, unlike most EEG experiments in which people wear an electrode cap, in this study, the electrodes had been implanted as treatment for chronic epilepsy. This allowed much more precise localization of the brain activity to specific regions.

Behaviorally, the researchers found that as the voice-feedback delay increased, people spoke more slowly. Even after taking this timing change into account, they found that increasing the delay led to greater activity in the superior temporal gyrus, a brain region used for processing sounds. While this region was expected to be part of the error signal, they also found corresponding activity increases in the upper region of the precentral gyrus—a brain area known for its role in controlling voluntary movements, but never before implicated in auditory feedback control.

“Dorsal precentral gyrus is a critical component of a cortical network that monitors auditory feedback to produce fluent speech,” Özker adds. “This region is engaged specifically when speech production is effortful during articulation of long utterances.”

#####

In your coverage, please use this URL to provide access to the freely available paper in PLOS Biology:   http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001493  

Citation: Özker M, Doyle W, Devinsky O, Flinker A (2022) A cortical network processes auditory error signals during human speech production to maintain fluency. PLoS Biol 20(2): e3001493. https://doi.org/10.1371/journal.pbio.3001493

Author Countries: United States

Funding: This study was supported by grants from the NIH (F32 DC018200 Ruth L. Kirschstein postdoctoral fellowship from the National Institute on Deafness and Other Communication Disorders to M.O. and R01NS109367 from the National Institute of Neurological Disorders and Stroke to A.F.) and the NSF (CRCNS 1912286 to A.F.) and by the Leon Levy Foundation Fellowship (to M.O.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.