What Your Music Playlist Lyrics Say About Your Intelligence
Researchers tracked what 185 people actually listened to on their smartphones for five months and found a subtle but measurable link between music choices and intelligence test scores. The surprise: lyrics mattered far more than the music itself.
Intelligence research tends to happen in sterile rooms with timed tests and anxious participants. A new study asked a different question: does cognitive ability quietly show up in something as casual as a music playlist? The early answer, drawn from real smartphone listening data, is a tentative yes, though the signal is modest and the research is still preliminary.
Song Lyrics, Not Beats, Predict Cognitive Ability
A new study published in the Journal of Intelligence tracked 185 participants over five months using a custom smartphone app, logging 58,247 unique songs and pulling detailed audio and lyrical data from databases including Spotify. The team, led by postdoctoral researcher Larissa Sust at Ludwig Maximilian University of Munich, then used machine learning models to see which of 215 musical features best predicted scores on a cognitive ability test covering fluid reasoning, vocabulary comprehension, and mathematical knowledge.
The headline result surprised even the researchers. Audio features such as tempo, key, and instrumentation contributed very little predictive power. Lyrics did most of the work. "When we looked more closely at how our prediction models worked and which aspects of music listening were most informative, one finding surprised us," Sust noted. "The lyrics of the songs people listened to were more useful for predicting cognitive ability than the musical features."
Which Lyrical Patterns Were Associated with Higher Scores
The lyrical associations are specific and worth unpacking. Participants whose playlists leaned toward songs about the present moment, perceived honesty, and home-related themes tended to score higher on cognitive tests. By contrast, playlists heavy on social words or tentative, hedging language predicted lower scores.
Emotional tone also played a role. People who preferred songs with less positive, more melancholic emotional content tended to have higher predicted intelligence scores. The researchers suggest this may reflect a tendency to use music for introspection and reflection rather than purely emotional uplift.
Two general listening habits added to the picture. Participants who spent more total time listening to music overall tended to score higher. And a preference for songs in languages other than German, the native language of the sample, was also associated with higher cognitive ability, possibly reflecting greater openness to unfamiliar linguistic and cultural content.
How the Study Captured Real-World Listening Behavior
Previous research linking music to intelligence typically relied on laboratory experiments or self-reported surveys, both of which introduce distortion. People misremember what they listen to, and lab settings change behavior. This study used passive digital tracking: the app quietly logged every song played on participants' smartphones across five months of ordinary life.
From the logged tracks, researchers gathered 215 features per participant, covering audio characteristics, lyrical content analyzed by a specialized linguistic tool that flags psychological themes and emotional tone, and general habits like total listening time and language preference. Two machine learning approaches were then tested: a linear LASSO regression and a nonlinear random forest model. The nonlinear model performed better, which the authors interpret as evidence that the relationship between music habits and cognitive ability is complex rather than a simple linear slope.
One notable feature that emerged as a strong predictor was "liveness," defined as the probability a track was recorded in front of a live audience. People who preferred low-liveness studio recordings tended to score higher on cognitive tests. The researchers speculate that higher-ability individuals may gravitate toward cleaner, more focused listening experiences.
Caveats: Correlation, Small Effect Sizes, and Unmeasured Confounders
The predictive power here is real but small. The study's authors are explicit that music listening alone is nowhere near sufficient for practical assessment of cognitive ability. The relationships are correlational, not causal: nothing in the data shows that listening to introspective lyrics raises intelligence, or that smart people deliberately curate cerebral playlists. Other factors not captured in the data, including age, education, and socioeconomic background, could be driving both music preferences and test scores simultaneously.
There is also a technical limitation specific to smartphone data. The app could not reliably distinguish between songs a participant actively chose and tracks that played automatically via algorithm or autoplay. That distinction matters for interpreting what a playlist reveals about deliberate taste versus passive exposure.
As Sust put it: "One key takeaway is that cognitive abilities may be reflected not only in tests or high-stakes performance but also subtly in everyday behavior." That is an intriguing possibility, but the study frames it carefully as a preliminary signal, not a diagnostic tool.
What Could Come Next in Digital Cognitive Profiling
The broader ambition behind this work is to build more ecologically valid ways of understanding cognition by reading the digital traces people leave in ordinary life. Combining music data with other behavioral streams, such as books read, places visited, or app usage patterns, could sharpen predictive accuracy considerably.
Researchers working on music's impact on brain plasticity and cognitive function have suggested that understanding these patterns could eventually inform healthcare and educational interventions. Separately, research from the University of California has explored how music-based approaches might be tailored to address specific cognitive challenges. The Munich team suggests their framework could, in the longer term, contribute to tools for adaptive digital environments or even early detection of cognitive decline, though that application remains speculative and would require far stronger predictive models than currently exist.
For now, the study offers a proof of concept: the songs you actually play, tracked passively over months, carry a faint but detectable imprint of how your mind works. Whether that signal is strong enough to be useful is a question for the next generation of studies.
Evidence rating
INVESTIGATIONAL — 69/100
Observational cohort study (n=185) using five months of smartphone-tracked music data and a brief mobile cognitive test. All findings are correlational; causal direction cannot be established. Potential confounders such as age are acknowledged by the authors and not yet controlled. Machine learning predictions explained only a small portion of variance in cognitive scores, and the study has not been independently replicated.
References
- Your music playlist might reveal subtle clues about your intelligence. https://www.psypost.org/your-music-playlist-might-reveal-subtle-clues-about-your-intelligence/
- Larissa Sust, Maximilian Bergmann, Markus Bühner, Ramona Schoedel (2026). Deep Beats, Deep Thoughts? Predicting General Cognitive Ability from Natural Music-Listening Behavior. Journal of Intelligence. https://doi.org/10.3390/jintelligence14020029
- Muriel T. Zaatar, Kenda Alhakim, Mohammad Enayeh, Ribal Tamer (2024). The transformative power of music: Insights into neuroplasticity, health, and disease. Brain, Behavior, & Immunity – Health. https://doi.org/10.1016/j.bbih.2023.100716
- Research shows that music may benefit our brains – UCnet – University of California. ucop.edu. https://www.ucop.edu/ucnet/news/2025/july/music-and-brains.html
