Syllable-timescale organization of auditory–motor feedback control
By integrating behavioral experiments, neurocomputational modeling, and EEG recordings, this study provides convergent evidence that auditory–motor feedback control in speech production is organized at the syllable timescale, where feedback delays disrupt speech and modulate neural suppression specifically when aligned with syllable duration.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Imagine you are trying to keep a perfect rhythm while tapping your fingers on a table. Now, imagine someone is recording your taps and playing them back into your ears, but with a tiny delay. If that delay is just right, your brain gets confused, your rhythm breaks, and you start tapping all over the place. This is the "Delayed Auditory Feedback" (DAF) effect, a trick scientists have used since the 1950s to study how we speak.
For decades, researchers noticed something weird: the delay that caused the most chaos was always around 200 ms (milliseconds). They guessed this happened because 200 ms is roughly the time it takes to say one syllable (like "ta" or "ba"). They thought the brain had a fixed "timer" that checked if the sound it heard matched the sound it expected to hear, and if the delay messed with that specific 200 ms window, the system crashed.
But here is the twist: What if the brain doesn't have a fixed timer at all? What if the brain's "timer" is flexible, stretching or shrinking to match exactly how fast you are talking?
That is exactly what M. Florencia Assaneo and her team at the Universidad Nacional Autónoma de México set out to find. They didn't just guess; they ran experiments, built computer brains, and even peeked inside the human brain with EEG sensors to solve the mystery.
The Race Against the Clock
First, the team put 15 people in a room and asked them to chant a sequence of syllables ("ta-te-ti-to-tu") at two different speeds: a slow, relaxed 3 syllables per second and a fast, frantic 6 syllables per second.
While they chanted, the researchers played their voices back with different delays: 30 ms, 80 ms, 160 ms, 240 ms, 320 ms, and 400 ms. They counted how many mistakes (like repeating a syllable or getting the order wrong) the speakers made.
The result was a "aha!" moment.
- When the speakers went fast (6 syllables/sec), the delay that caused the most chaos was around 80 ms.
- When the speakers went slow (3 syllables/sec), the chaos peaked at 160 ms.
The "most disruptive" delay didn't stay fixed at 200 ms. Instead, it scaled perfectly with the speed of the speech. If you speak twice as fast, the brain's "sweet spot" for getting confused happens twice as quickly. This suggests the brain isn't using a rigid, pre-set clock. Instead, it seems to be using a flexible window that matches the length of the syllable you are currently saying.
The Computer Brain Test
To make sure this wasn't just a fluke, the researchers built three different "computer brains" (neurocomputational models) to see which one could explain the messy human data.
- The "Point-by-Point" Robot: This model compared the sound wave at every single tiny moment. It failed to reproduce the human results.
- The "Rhythm" Robot: This model only cared about the timing and rhythm, ignoring the actual sound content. It also failed.
- The "Syllable" Robot: This model acted like a human. It didn't check every millisecond. Instead, it waited until a whole syllable was finished, compressed that chunk of sound into a simple summary, and then compared it to what it expected.
Only the Syllable Robot successfully recreated the experiment's results. It showed that the brain likely groups sounds into syllable-sized chunks before checking them. If the delay messes up the timing of that specific chunk, the brain panics. This simulation suggests that the "syllable" is a fundamental unit of how we control our speech, acting like a natural container for our auditory feedback.
Listening to the Brain's "Shh!"
Finally, the team wanted to see what was happening inside the brain. They used EEG (electroencephalography) to record electrical activity from 28 participants.
When we speak, our brain usually sends a "shh!" signal to our ears, dampening the sound of our own voice. This is called Speech-Induced Auditory Suppression (SIAS). It's like the brain saying, "I know I'm making that noise, so I don't need to listen to it as loudly."
The researchers found that this "shh!" signal depends on the syllable speed, too. When the feedback delay was fixed at 250 ms, the brain's suppression was stronger when people spoke slowly (3 Hz) and weaker when they spoke quickly (6 Hz).
This confirms that the brain's ability to predict and dampen its own voice isn't just about the absolute time delay. It's about how that delay fits into the rhythm of the syllable. If the delay breaks the rhythm of the syllable, the brain's prediction fails, and the "shh!" signal gets weaker.
The Verdict
So, what did we learn?
- The Main Finding: The brain organizes speech feedback around the syllable. It uses a flexible window that stretches or shrinks to match how fast you are talking, rather than a fixed 200 ms timer.
- What Was Ruled Out: The idea that the brain uses a simple, continuous comparison of sound waves at every millisecond, or that it relies solely on a fixed rhythm that doesn't change with speaking speed. The data and simulations show these simpler ideas don't work.
- How Sure Are We? The team has convergent evidence. They have measured this behavior in humans, simulated it in computer models that only the syllable-based one could replicate, and observed the corresponding neural signals in the brain.
While the paper doesn't claim to have solved the entire mystery of speech, it provides strong, direct evidence that the syllable is a key building block in the brain's speech-control architecture. It's as if the brain speaks in "chunks," and if you mess up the timing of a chunk, the whole conversation stumbles.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.