Summary: In a novel approach to psychiatric research, a Yale study indicates that songwriting and group music-making can help people with psychosis reconnect with reality. The research frames music as a way to engage the brain’s predictive coding system — its capacity to anticipate what comes next — and shows how rhythm and melody can restore a sense of shared experience and reduce social withdrawal.
Participants living with schizophrenia and frequent auditory hallucinations reported lower levels of paranoia and clearer social engagement after taking part in weekly two-hour group songwriting sessions, suggesting music may be a powerful therapeutic tool alongside existing treatments.
Key Findings:
- Reduced paranoia: Although hallucinations did not stop for all participants, those with less severe auditory symptoms experienced a measurable decrease in paranoid thoughts following the songwriting workshops.
- Shift from “I” to “We”: Linguistic analysis showed participants used fewer first-person singular pronouns (I, me, mine) and more plural pronouns (we, us, ours) after the sessions, signaling a return to social connection and shared identity.
- Fewer negative side effects than medication: Unlike some antipsychotic medications that can cause lethargy or cognitive dulling, the music-making sessions produced no reported adverse effects; participants appeared more engaged, expressive, and emotionally present.
- Potential lasting brain changes: The lead researcher, Dr. Philip Corlett, proposes that repeated musical engagement may retrain the brain’s prediction systems in a lasting way, offering a neuroplastic route to recovery.
Source: Yale
Our brains predict incoming sensory signals—sights, sounds, tastes, smells, and touch—based on past experience. When you bite an apple, for example, you expect a certain crunch and sweetness because similar events in the past set that expectation.
This process, often called predictive coding, reduces cognitive load and helps learning. But when predictions become too strong or decoupled from external reality, they can contribute to hallucinations and delusions that characterize psychosis, a condition where perception and belief diverge from shared reality.
Published April 9 in the journal Psychosis, the new study from Yale explores whether structured group music-making can help people experiencing psychosis re-engage with their environment by exercising the brain’s predictive abilities.
“Music is a golden road for making predictions,” says Philip Corlett, PhD, associate professor of psychiatry at Yale School of Medicine and senior author of the study. A familiar lyric—“Sweet Caroline…”—automatically brings the expected melody to mind. That immediate, reliable call-and-response is precisely what the research team sought to harness.
Because music strongly links expectation and confirmation, Corlett’s Belief, Learning, and Memory Lab collaborated with community music facilitators to test whether songwriting could help stabilize perception in people experiencing hallucinations.
“People may have hallucinations because their internal predictions overwhelm incoming sensory evidence,” Corlett explains. “Music provides a structured, safe way to violate and then confirm expectations, like a therapeutic exercise for the brain’s predictive machinery.”
Corlett partnered with Adam Christoferson, a music facilitator and founder of Musical Intervention, a New Haven organization that uses collective music-making to support rehabilitation and community connection. Corlett’s interest in music therapy grew from clinical observations: patients who participated in music sessions often appeared more animated, expressive, and socially connected than in other settings.
Group music eases paranoia
For the study, the team recruited 20 people aged 18 to 65 who either had schizophrenia or experienced weekly distressing auditory hallucinations. Over six weeks, participants completed baseline psychometric assessments for hallucinations and paranoia, then joined small groups of five for four weekly two-hour songwriting and recording sessions led by a professional musician. Each group received basic recording equipment—microphone, guitar, keyboard, and drums—and participants were encouraged to write original lyrics.
After the final session, participants repeated the same assessments and took part in a post-session interview. The research combined standard symptom questionnaires with objective linguistic analysis using Linguistic Inquiry and Word Count (LIWC) to track changes in language tied to mental states.
While hallucination frequency did not fall for all participants, those with milder symptoms reported less paranoia. Linguistic results showed a notable decline in first-person singular pronoun use and increases in plural pronouns and language related to agency, accomplishment, cognitive processing, and positive emotion—suggesting stronger social identity and improved mood.
Group music-making delivered benefits beyond symptom measures: participants reported decreased isolation, renewed creativity, and a sense of belonging. For facilitators like Christoferson, the findings confirmed years of field experience that music fosters identity, emotional expression, and practical skill-building for people in recovery.
“This work shows we can evaluate music-based interventions with rigorous clinical research,” Corlett says. “Music therapy may not replace medication, but it can complement clinical care, reduce social isolation, and engage patients in ways medications sometimes cannot—without the negative side effects.”
Corlett and his team are now investigating how repeated music-making might change brain circuitry. “I suspect consistent musical engagement produces lasting neural changes,” he says, “and our next step is to identify the mechanisms behind those changes.”
Key Questions Answered:
A: The familiar melody creates an immediate prediction and confirmation loop. For minds that struggle to separate internal noise from outside signals, this rhythmic and melodic expectation functions like cognitive exercise, practicing the brain’s ability to match prediction to reality.
A: Not at present. The study describes music-making as a powerful non-clinical tool that reduces social isolation and stimulates creativity without medication side effects. It is best seen as a complementary approach that treats the whole person, not only symptoms.
A: Language reveals social and emotional states. Increased use of “we” reflects a shift out of isolation and toward shared identity and belonging—an important marker of recovery for people whose psychosis can create a deeply solitary experience.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by staff to clarify research methods and outcomes.
About this music and mental health research news
Author: Colleen Moriarty
Source: Yale
Contact: Colleen Moriarty – Yale
Image: The image is credited to Neuroscience News
Original Research: Closed access.
“Song-making in a group (SING): a longitudinal study for people experiencing psychosis” by Deanna L. Greco et al., published in Psychosis.
DOI: 10.1080/17522439.2026.2634654
Abstract
Song-making in a group (SING): a longitudinal study for people experiencing psychosis
Background
Creative expression turns imagination into shared experience. Music-making mobilizes perception in ways both creators and listeners can feel, and collaborating with others amplifies an individual’s voice within a group. This collective creative process can foster agency and belonging—qualities that may be especially beneficial to people who experience isolation or alienation. This study evaluates how group song-making affects people living with psychosis.
Methods
Twenty participants with psychosis took part in four collaborative sessions where they wrote and recorded songs with four peers and a music facilitator. The study combined standard symptom measures for paranoia and hallucinations with objective linguistic analysis to measure changes in mental states. Symptom questionnaires were administered before and after the music intervention, while shifts in language were captured using Linguistic Inquiry and Word Count (LIWC) 2022.
Results
Although overall hallucination frequency did not decline across the group, paranoia decreased among those with less severe hallucinations. Linguistic analysis showed a significant reduction in first-person singular pronoun use and increased use of plural pronouns, along with more language reflecting accomplishment, agency, cognitive processing, and positive emotion.
Discussion
The observed language shifts are discussed in relation to how song-making might influence participants’ lives beyond the workshop. Overall, the study highlights group-based songwriting as a promising intervention to support recovery from psychosis by promoting social connection, emotional expression, and cognitive engagement.