san jose collage
San Jose
Serving San Jose, California (Santa Clara County)Vol. 1, No. 25
Medicine

The brain’s ‘listening’ center also plans every word you say, UCSF finds, and that could help people who have lost their speech

A 150-year-old textbook idea gets rewritten:  a small patch of Wernicke’s area builds a plan for each word a quarter-second before it is spoken.

ShareFacebookXEmail
Diagram of the left side of a human brain with two outlined regions labeled Broca’s area, toward the front, and Wernicke’s area, toward the back, joined by a curved bundle of lines.
A textbook diagram of the brain’s two classic language areas, Broca’s area and Wernicke’s area. — Illustration:  OpenStax College, via Wikimedia Commons (CC BY 3.0)

A part of the brain that textbooks describe as a listening center also writes the plan for each word a person is about to say, scientists at UC San Francisco reported Nov. 19 in the journal Science.  The finding could help surgeons protect speech during brain operations and could lead to brain implants that give a voice back to more people who have lost it, according to UCSF.

The study comes from the lab of Dr. Edward Chang, a UCSF neurosurgeon whose team builds brain-computer interfaces, devices that read brain signals and turn them into words.  “This changes the textbook view of how the brain perceives and produces speech,” said Chang, the study’s senior author.

What this means

For people who cannot speak after a stroke, brain injury or paralysis, this points to a new place to read the words they are trying to say.  Today’s speech implants decode signals from the brain’s movement areas, which steer the lips, tongue and voice box.  Those areas can be damaged in some patients.  The newly mapped region works one step earlier, so it could offer another route, UCSF said.

The team showed the idea can work in a person with severe paralysis who could no longer speak:  from this region alone, they decoded words the person tried to say and speech the person heard.  That was a single-person demonstration, not a treatment.  The map could also help brain surgeons know more precisely which spots to avoid so a patient can still talk after an operation.

A 150-year-old puzzle

In the 1870s, German neurologist Carl Wernicke noticed that patients with damage toward the back of the brain’s temporal lobe, on its side, had trouble both understanding speech and producing it.  Because that region sits inside the auditory cortex, which handles sound, it came to be taught mainly as the place where we understand speech.  Speaking was credited to other areas, chiefly Broca’s area toward the front of the brain.

That left a puzzle:  Why do strokes in a “listening” area also leave people struggling to say the words they want?  The new study answers it.  A small region called the left posterior superior temporal gyrus, or lpSTG, turns on about a quarter-second before a person says a word, according to UCSF.

What the brain’s word plan contains

The activity is not about the word’s meaning.  It is a blueprint holding the word’s length, its number of syllables and how common it is.  Brain areas that control the mouth then turn that plan into the movements needed to say it.  The region tracks whole words rather than single sounds or mouth movements, the paper reports.

Neurons that respond when we hear speech sit right beside the planning neurons, and sometimes overlap with them.  “One of the most important skills we have as infants is learning to turn the words we hear into words we can say,” said Matthew Leonard, a UCSF associate professor of neurological surgery and the study’s co-first author.  Having one area do both jobs may be key to how children learn language, he said.

How the scientists tested it

Leonard and co-lead author Alex Silva, a student in UCSF’s combined M.D.-Ph.D. program, combined 15 years of data from Chang’s lab with new experiments.  They used recordings taken directly from patients’ brains, down to single neurons, plus mild electrical stimulation and studies of people with damage in the region, according to the paper.  Briefly disrupting the lpSTG just before a person spoke caused speech errors, evidence that the region is needed to produce words, not just to listen.

Source:  Silva et al., Science, Nov. 19, 2026; UCSF news release via Medical Xpress.

Drafted by AI from the linked public source. Approved by a Daily News Editor.

Sources and documents