A new study from the University of California, San Francisco has upended a long-held assumption in neuroscience about how the brain organizes the sounds of speech. For over 150 years, Broca’s area—a region in the frontal lobe—was believed to be the main hub responsible for translating thoughts into the coordinated movements required for speaking. But new research published in Nature Human Behaviour provides evidence that another area, the middle precentral gyrus, plays a far more central role in this process than previously recognized.
The study found that this region of the brain not only helps sequence speech sounds into coherent utterances but also plays a direct role in orchestrating the motor plans necessary to produce them. The discovery reshapes our understanding of how fluent speech is generated and may lead to better tools for treating speech disorders or developing brain-computer interfaces for people who cannot speak.
The researchers conducted the study to better understand how the brain plans and sequences the movements needed for speech—a process known as speech-motor sequencing. For decades, scientists believed this function was housed primarily in Broca’s area, a region in the frontal lobe historically linked to language production. But clinical observations and new neuroimaging findings have begun to cast doubt on this assumption.
In particular, some patients with damage to Broca’s area retain normal speech abilities, while others with injuries to neighboring areas, including the middle precentral gyrus (mPrCG), show marked speech planning difficulties. This discrepancy prompted the researchers to reexamine the brain mechanisms behind fluent speech.
Lead investigator Edward Chang, a neurosurgeon and professor at the University of California, San Francisco, has spent years studying how the brain produces spoken language. His interest in the mPrCG was sparked by a case in which a patient developed apraxia of speech—a condition that impairs the ability to coordinate speech movements—after surgery that removed part of this region. The same speech impairment did not occur in other patients whose surgery involved Broca’s area.
Jessie Liu, then a graduate student in Chang’s lab, played a key role in shaping the research. While developing neural interface devices to help people with paralysis communicate, Liu and Chang noticed consistent speech-related activity coming from the mPrCG. These observations led them to suspect that the region might do more than control pitch, as previously thought.
“The field was making a lot of progress in studying articulation, that is the low-level movements that control the vocal tract for speech, so we started to wonder about the neural processes that occurred prior to articulation, which the field knew much less about,” explained Liu, who is now a postdoctoral scholar.
“Generally, that would be called speech planning, and we became particularly interested in a subprocess of that called speech-motor sequencing as there are speech disorders where just that specific process is malfunctioning. Overall, speech-motor sequencing was something that has important implications for both basic and clinical sciences, and that really excited us.”
The researchers hypothesized that the mPrCG could be responsible for ordering the individual speech sounds—syllables and phonemes—into coherent sequences, acting as a bridge between abstract speech planning and motor execution. The study was designed to test this hypothesis directly by recording and stimulating brain activity in people as they produced spoken sequences of varying complexity.
To investigate how the brain sequences the movements required for speech, the researchers recruited 14 English-speaking volunteers who were undergoing surgery for epilepsy. As part of their medical treatment, each patient had been implanted with high-density electrocorticography (ECoG) grids on the surface of their brain. These grids allowed the researchers to record neural activity directly from multiple brain areas while the participants performed speech tasks. All participants had normal speech and motor abilities and no history of speech disorders, making them ideal candidates for studying typical speech production.
Participants were asked to perform a delayed speech repetition task. During each trial, they were shown a sequence of syllables on a screen—such as “ba-ba-ba,” “ba-da-ga,” or more complex combinations like “blaa-draa-gloo”—and after a brief delay of about one second, they were cued to repeat the sequence out loud. This design allowed the researchers to separate the brain activity related to seeing the syllables, holding them in memory, preparing to speak, and finally producing the sounds.
By varying the complexity of the sequences, the researchers could measure how the brain responds to different speech planning demands. Some sequences were simple, using the same syllable repeated three times, while others included three different syllables. Some also included more complex articulatory features, like consonant clusters, which are harder to pronounce.
Using the ECoG recordings, the researchers tracked neural activity across multiple areas of the brain, focusing on how it changed during the four phases of the task: visual encoding, memory delay, preparation to speak, and actual speech production. They found that many brain regions showed brief bursts of activity during one or two of these stages.
But a particular set of electrodes, especially those placed in the middle precentral gyrus (mPrCG), displayed sustained activity that lasted through all four stages of the task. These “sustained” signals suggested that the mPrCG was continuously involved in speech planning from the moment participants saw the syllables until they finished saying them.
The researchers noticed that the mPrCG stood out in several ways. Not only did it contain more sustained electrodes than any other region, but the activity within it also varied in meaningful patterns. Using unsupervised clustering techniques, the team identified four distinct temporal profiles of sustained activity. Some electrodes showed stronger responses during the initial reading of the syllables, while others peaked closer to the moment of speech.
The mPrCG had a relatively even mix of these patterns, suggesting it plays a flexible and ongoing role throughout the speech preparation process. This contrasted with other brain regions, like Broca’s area, which showed less involvement across all task phases.
To test whether the mPrCG was specifically involved in sequencing, the researchers examined whether its activity changed depending on the complexity of the syllable sequences. They found that the region was especially sensitive to “sequence complexity”—whether the syllables were repeated or varied—not just to how difficult they were to pronounce.
In contrast, regions like the ventral precentral gyrus and supramarginal gyrus were more responsive to articulatory complexity, meaning they were activated by sounds that were physically harder to produce. The mPrCG’s response to sequence structure—not just sound difficulty—supported the idea that it helps organize speech sounds into a meaningful order before they are spoken.
The researchers also analyzed whether activity in the mPrCG predicted how quickly a person began to speak. They found that certain electrodes in the mPrCG could reliably forecast the participant’s reaction time—the interval between the go-cue and the start of speech—on a trial-by-trial basis. This finding linked mPrCG activity directly to motor planning, showing that this region does more than just monitor or respond to speech; it helps initiate it. Electrodes that predicted reaction time often overlapped with those that were sensitive to sequence complexity, strengthening the evidence that the mPrCG coordinates the timing and order of speech movements.
“We had hypothesized that the mPrCG was involved in speech-motor sequencing but it was surprising just how robust the activity was,” Liu told PsyPost. “The use of ECoG (electrocorticography) allowed us to have very high spatial and temporal resolution, so we could compare activity across areas at 5 ms resolution. This made it very clear that the mPrCG had strong effects of sequence complexity across the whole task.”
In a key part of the study, the researchers applied electrical stimulation to the mPrCG in five participants during the speech task. This type of stimulation can temporarily disrupt normal brain activity, allowing scientists to see what happens when a specific area is impaired. When the mPrCG was stimulated, participants made speech errors that resembled symptoms of apraxia of speech—a disorder in which people struggle to coordinate their speech despite knowing what they want to say.
These errors were most common when participants tried to say complex sequences. They included stretching out syllables, inserting unintended pauses, mispronouncing sounds, and stuttering. Notably, these errors did not occur during simple sequences or during tasks that involved making mouth movements without speech, suggesting the effects were specific to sequencing and not general motor control.
Importantly, stimulation of Broca’s area did not cause these types of errors, despite its historical association with speech planning. This contrast suggested that Broca’s area might play a more abstract or supervisory role in language, while the mPrCG directly handles the motor sequencing necessary to turn speech plans into coordinated muscle movements.
Together, the results revealed that speech-motor sequencing is not confined to Broca’s area, but instead relies on a distributed network of brain regions. Within this network, the mPrCG plays a central role, serving as a hub where phonological plans are translated into motor instructions for speaking. This challenges long-standing models of speech production and may reshape how neuroscientists understand language in the brain.
“Speech-motor sequencing is a necessary process that allows us to flexibly build words from smaller units like syllables,” Liu explained. “This was thought to occur in a part of the brain called Broca’s area, but this has become controversial, as studies also show that damage to Broca’s area does not result in speech-motor sequencing errors. To test this, we recorded brain activity from Broca’s area and many other parts of the brain that are related to speaking while participants spoke different speech sequences.”
“Instead of Broca’s area or any other area we recorded from, we found that an area called the middle precentral gyrus (mPrCG) had brain activity crucial for speech-motor sequencing. This has important impacts on basic science as we show that the mPrCG should be included in future models of speech production. Additionally, this will help surgeons make more complete maps of important speech areas in a patient’s brain, so that they can carry out necessary surgery while still preserving a patient’s ability to speak fluently.”
While the study provides evidence for the role of the mPrCG in speech-motor sequencing, it has some limitations. The participants were all epilepsy patients undergoing surgery, which means their brains may not represent the general population. Grid placement was based on clinical need, so not all brain regions were sampled evenly.
“Though we recorded from many brain areas, these were all on the surface of the cortex,” Liu noted. “Subcortical regions also play roles in speech production and motor control, so testing those is an important goal for future studies.”
Still, the study opens new avenues for exploring how the brain coordinates complex behaviors like speech. Future research may investigate how the mPrCG interacts with other brain regions involved in language, whether its role extends to other types of sequencing like reading and writing, and how it could be targeted in rehabilitation or assistive technology.
“Our work in this paper is really just the first step concretely showing the role of the mPrCG in sequencing,” Liu told PsyPost. “Ultimately, we want to understand the neural computations and exact mechanisms of the mPrCG in speech-motor sequencing, as well as how this region interacts with the broader network of speech areas. Improved recording techniques, like recording from single neurons, as well as new methods to model these neural populations will be critical for achieving this.”
“This work would not have been possible without the generosity of the patient volunteers who donated their time and effort to this study,” Liu added. “This work was also a strong team effort with my co-first author Lingyun Zhao, PhD.”
The study, “Speech sequencing in the human precentral gyrus,” was authored by Jessie R. Liu, Lingyun Zhao, Patrick W. Hullett, and Edward F. Chang.
