Election Results 2026 MidtermsView →
Dock Line
The Sarah M. and Charles E. Seay Building, a tan brick academic building with a wide overhanging roof, on the University of Texas at Austin campus, photographed Aug. 3, 2019
The Sarah M. and Charles E. Seay Building on the University of Texas at Austin campus, photographed Aug. 3, 2019. A Nov. 16 university release describes a psychology study, published Nov. 12 in PLOS Biology, on how the brain learns which sounds to amplify. (Michael Barera)

Jane Long

November 17, 2026 · 4 min read

UT Austin Study Finds the Brain Learns Which Sounds to Amplify, Even Without Trying

A Nov. 16 University of Texas at Austin release describes a study, published Nov. 12 in PLOS Biology, finding that the brain learns which sounds have proved important and amplifies them even without deliberate effort.

A University of Texas at Austin news release dated Nov. 16 describes a study finding that the brain learns which sounds have proved important and amplifies them without any deliberate effort to listen. The paper itself was published Nov. 12 in PLOS Biology (DOI 10.1371/journal.pbio.3004076). Dock Line read both the university release and the journal article.

"You can hear your name across a noisy room even when you weren't listening for it," the release says, introducing the everyday example the researchers used to frame the work. Lori Holt, professor and chair of UT's Psychology Department and a co-author of the paper, said in the release, "Our brains learn to recognize important sounds like speech or a spouse's voice." For the same sound, she said, the brain can "turn up the dial on the specific cues that convey who is talking versus what they are saying, depending on the task at hand."

A made-up world of sounds, watched by a scanner

The paper, by Sahil Luthra of Stony Brook University, Raha N. Razin of University College London, Adam T. Tierney of Birkbeck, University of London, Lori L. Holt at UT Austin, and Frederic Dick of University College London, reports two auditory category-learning tasks. Listeners had to make sense of a completely unfamiliar "soundscape" of chirps, then sort those sounds by the identities and sizes of cartoon "aliens."

According to the article, 95 young adults trained for five days, with feedback, to categorize the novel sounds. In the main task, four sound categories were paired with four alien identities. Each sound carried energy in two non-overlapping frequency bands, one roughly 80 to 750 hertz and the other roughly 1,000 to 9,000 hertz. Only one of those bands contained the pattern that actually identified the alien. The other band had varying chirps that did not predict identity. A second task asked listeners to judge alien size from amplitude, a different cue.

Of the 95 people in the identity-and-size training group, 54 reached more than 75% accuracy by day 5 and were recruited as "categorization experts" for a functional MRI session. Five of those were omitted for motion or noncompliance, leaving 49 people in the final scanner group. A separate group of 20 people trained only on the size judgment, with no identity training, and also completed an MRI session. Those figures come from the journal article, not the campus release.

What the scanner showed

The paper reports that only participants who had become expert at alien-identity categorization showed a neural prioritization of the frequency band that carried the identity cue. That bias persisted even when the in-scanner task shifted to size categorization, for which the identity band was no longer useful. People who had heard similar sounds without learning to tell the aliens apart did not show the same shift, the article says. The university release puts the same point more plainly: the change came from learning, not just from hearing.

The researchers also compared people who mastered the identity task quickly with people who improved more slowly over the five days, and they compared higher and lower performers on a generalization test. Both the paper and the release say learners can reach a similar level of skill along different paths, and that those paths leave different patterns of brain activity. The release says that may help explain why people without hearing or language problems still differ in how they process speech. The paper makes a similar point about individual differences that persist even in adults without those difficulties.

Holt said in the release, "The brain doesn't simply hear the world as it is. Every day, experience teaches the brain what to amplify and what to tune out. Practice doesn't just make us better at using information. It changes which information reaches the front of the line."

What the study does not prove

The university release says the findings "may help explain the auditory-learning difficulties linked to communication disorders such as dyslexia." The paper is more careful. In the discussion, the authors say the work has "implications for understanding and potentially remediating learning-related communication disorders like dyslexia that have been linked to auditory learning." They did not test people with dyslexia, and they did not show that the training repairs a disorder. That is a possible future use, not a result of this experiment.

The authors also note a limitation in a related attention task used for comparison: the task was difficult, especially for some participants, and it was run at the end of a long MRI session, when some listeners were fatigued. The work was supported by grants from the National Institutes of Health and the National Science Foundation, and by an individual fellowship from the National Institute on Deafness and Other Communication Disorders, the paper says. The authors reported no competing interests.

The study is a laboratory finding about how training can change which sounds the brain favors. It is not a clinical trial, and it does not tell a parent or a teacher what to do tomorrow. The honest claim is smaller, and still worth reporting: once people learned which part of a sound carried the useful information, their brains kept favoring that part even when the task changed.