During the COVID-19 pandemic, many people reported turning to music not just for entertainment, but for comfort, support, and even companionship. Now, a new study published in Scientific Reports provides evidence that this sense of “music as company” may be more than a metaphor. Researchers found that music listening can shape mental imagery by increasing the presence of social themes in people’s imagined scenes.
The idea that music offers social comfort has been widely reported in surveys and interviews, especially during periods of isolation such as pandemic lockdowns. Listeners often say they use music “to keep them company” or to ease feelings of loneliness. But the extent to which music genuinely prompts social thinking—rather than simply modulating mood—has been unclear. Most prior research has focused on how music affects memory, emotion, or passive mind-wandering. Few studies have examined how music shapes the content of intentional mental imagery, particularly whether it elicits social scenes or interactions.
This distinction is important because directed mental imagery is used in various clinical and therapeutic settings. Techniques such as imagery rescripting or exposure therapy rely on a person’s ability to vividly imagine scenarios. If music can reliably shift the content of such imagery toward social themes, it might offer new ways to enhance therapeutic outcomes or support individuals struggling with loneliness.
“There have been many reports of people listening to music to ‘keep them company,'” said study author Steffen A. Herff, a Horizon Fellow and leader of the Sydney, Music, Mind, and Body lab at the Sydney Conservatorium of Music at the University of Sydney. “The number of these reports was particularly high during the pandemic isolation periods. But whether this is just a figure of speech, or an actual empirically observable effect of music on social thought was previously unclear, despite its great applicational implications.”
To explore this, the researchers designed two experiments involving over 600 participants. In the first experiment, participants were asked to perform a directed imagery task. They watched a brief video clip showing a solitary figure beginning a journey toward a distant mountain, and were then instructed to close their eyes and imagine how the journey continued. During this 90-second imagination phase, they either heard no sound or listened to folk music in Spanish, Italian, or Swedish.
Each song was played in both vocal and instrumental versions, and participants were either fluent or non-fluent in the language of the lyrics. This allowed the researchers to test whether comprehension or vocal presence mattered for the effect. Across the three language groups, 600 participants took part, split evenly between native and non-native speakers of each language.
After each imagination trial, participants described what they imagined and rated aspects such as vividness and emotional tone. These descriptions were then analyzed using a topic modeling technique called Latent Dirichlet Allocation, which allowed the researchers to identify recurring themes across participants’ narratives.
The researchers found strong evidence that music had an impact on the characteristics of mental imagery. Compared to silence, music consistently led to more vivid mental scenes, more positive emotional tone, and greater perceived time and distance traveled in the imagined journey.
More notably, music also increased the presence of social themes. One of the nine identified topics—labeled Topic I—was clearly centered around social interaction, including words like “people,” “friend,” “village,” and “together.” This topic appeared far more frequently in participants’ imagery when they were listening to music. The effect held across nearly all music conditions tested (30 out of 36), suggesting a consistent influence of music on the presence of social content in imagined scenes.
“The effect of induced social interactions into imagination was much stronger than we originally anticipated,” Herff told PsyPost. “The probability of imagination to contain social interactions in our experiment is more than three times higher when participants listen to music, compared to silence.”
This shift toward social imagery occurred regardless of whether the song included lyrics or whether the listener understood the language. The effect also remained consistent when the vocals were removed, suggesting that the presence of a human voice or semantic content was not necessary for the social effect to emerge.
“Music’s ability to increase social imagination works even if you don’t understand the lyrics of the song, for example because it is in a different language,” Herff said. “In fact, it even works if there are no lyrics at all! Together, this tells us that it’s not simply a question of hearing the human voice that is driving this.”
One exception occurred with an Italian folk song describing a communal grape harvest, where understanding the lyrics amplified the effect—highlighting how specific lyrical content can enhance music’s social influence under certain conditions.
In a second experiment, the researchers used a stable diffusion model to generate images based on participants’ written descriptions of their mental imagery. These visualizations allowed for a more intuitive grasp of the differences between imagery during music and silence.
A new group of 60 participants then viewed pairs of these images—one generated from a music condition, the other from silence—and tried to guess which image was imagined while listening to music. Half of these participants completed the task in silence, while the other half listened to the same music that the original participant had heard.
Those who listened to music during the task performed better, suggesting that music provides contextual cues that help people interpret others’ imagined content. In effect, they were better able to recognize the emotional or thematic signature of music-influenced imagery when they were themselves immersed in the same auditory context.
“Interestingly, when a new group of participants was provided with representations of what the initial participants imagined during silence and during music, they could tell which content was previously imagined during music listening, and which was imagined during silence, but only if the new participants also listened to the music,” Herff told PsyPost. “This tells us that there is a ’theory of mind’ when it comes to music-evoked mental imagery. In other words, you can imagine what someone else might imagine when listening to music.”
To further validate the results, a research assistant unaware of the study’s design manually annotated all 4,200 participant responses for signs of social interaction, temperature, brightness, and narrative perspective.
This analysis showed that descriptions written after music trials were more likely to include social elements and described warmer, brighter environments. In contrast, descriptions following silence tended to be darker, colder, and lonelier. About 39% of music condition descriptions included some form of social interaction, compared to just 12% in the silent condition.
These findings add another layer of evidence that music listening can facilitate social thought, even when people are engaging in a solitary and abstract task like mental imagery.
“Mental imagery is a notoriously difficult phenomenon to study due to its elusive and deeply personal nature,” Herff said. “With our directed mental imagery paradigm, we can now collect data on what people imagine with enough constraint (i.e., directed imagined journey from a clearly defined starting point towards a clearly defined topographical landmark) to do comparisons across conditions (e.g., music vs silence) but also enough freedom (i.e., anything could happen during these imagined journeys) for people to truly unfold their imagination.”
“I believe our findings provide support for an intuition about imagination, music, and their interaction, that many who explore the topic already have, no matter if they approach it from an empirical, artistic, or philosophical perspective. But where previously we had to rely on our intuition, we now have something more tangible to build upon.”
Although the findings were consistent across multiple songs and languages, the study’s musical selections were limited to the folk genre and to Western cultural contexts. Future work is needed to explore whether these effects generalize to other types of music—such as pop, jazz, or electronic—and to non-Western musical traditions.
“Ideally, we would have tested a much larger and more diverse set of music, in particular non-western music, and for each of them, included an expert familiar with that given music and culture,” Herff noted. “However, further increasing the stimulus set and number of recruited participants would have made this already logistically challenging endeavour unfeasible. But that is certainly something we have our eyes on for the future.”
It also remains unclear what specific musical features drive the effect. Is it melody, rhythm, tempo, or cultural associations that make a piece of music more likely to elicit social thought? Answering these questions would help refine music-based interventions in clinical and therapeutic settings.
“This study is part of a larger scale investigation into music-evoked mental imagery,” Herff said. “Currently we are investigation music-evoked mental imagery by looking both at very detailed musical features and how they shape mental imagery. For example, we just published this study: Micro-variations in timing and loudness affect music-evoked mental imagery. We are also looking closer at how listeners use music systematically to self-regulate. For example, in this recently published study, we explored how older adults engage with music when they feel lonely: Music as social surrogate? A qualitative analysis of older adults’ choices of music to alleviate loneliness.”
“At the same time, we are working closely together with the music community to understand the insights and intuitions on how to use music to shape listeners’ imagination that already exists in these experts. We hope that our research can contribute to clinical (e.g., cognitive behaviour therapies that use mental imagery techniques), recreational (e.g., roleplay), and artistic applications (e.g., new compositions).”
The study, “Solitary silence and social sounds: music can influence mental imagery, inducing thoughts of social interactions,” was authored by Steffen A. Herff, Gabriele Cecchetti, Petter Ericson, and Estefania Cano.
